Skip to main content

Suharjito MT

Bina Nusantara University, MTI, Faculty Member

Followers

272

Following

132

Co-authors

11

Public Views

Suharjito is the Head of Information Technology Department in Binus Graduate Program of Binus University. He received under graduated degree in mathematics from The Faculty of Mathematics and Natural Science, GadjahMada University, Yogyakarta, Indonesia in 1994. He received master degree in information technology engineering from Sepuluh November Institute of Technology, Surabaya, Indonesia in 2000. He received the PhD degree in system engineering from the Bogor Agricultural University (IPB), Bogor, Indonesia in 2011.His interest research are intelligent system, Fuzzy system, image processing and software engineering.

less

University of Libya

TJPRC Publication

Michael Alexander Kaufmann

University of Fribourg

Leander Schietgat

KU Leuven

Moustafa Ghanem

Imperial College London

Nik Nailah Binti Abdullah

Monash University (Malaysia)

Magdalini Eirinaki

San Jose State University

Ijariit Journal

Interests

Uploads

Papers by Suharjito MT

Adaptive Neuro-Fuzzy Inference System for Texture Image Classification

—One of the most important problems in pattern recognition is texture-based image classification.... more —One of the most important problems in pattern recognition is texture-based image classification. In this paper , the combination of Discrete Cosine Transform (DCT) and Gray Level Co-Occurrence Matrix (GLCM) methods for feature extraction was proposed. The attributes extracted from DCT method were mean and variance, while the attributes extracted from GLCM method were energy and entropy. Adaptive Neuro-Fuzzy Inference System (ANFIS) was used as a classifier. The classifier model was trained using 50% of the texture images and remaining images were used for testing. Four classes of texture images were downloaded from KTH-TIPS (Textures under varying Illumination, Pose and Scale) image database, three of which were used in each experiments thus there were four data combination. The best data testing accuracy result towards textures of crumpled aluminium foil, corduroy, and orange peel is 98.3%, which is 1.6% better than one hidden-layer feed forward neural network classifier. In average, testing accuracy result of ANFIS excelled one hidden-layer feed forward neural network with 93.7% over 90.4%.

Human Ear Recognition Methods Based on Image Rotation

– The ear is part of biometrics that has a unique and stable structure, avoided from aging. This ... more – The ear is part of biometrics that has a unique and stable structure, avoided from aging. This study aims to look at the accuracy of Geometric Moment invariant (GMI) and Zernike Moment invariant (ZMI) with Self-Organizing Maps to recognize human ears. This study used data from ear AMI database, consisting of 125 ears from 25 different people, in which each person has 5 images of the right ear. In this study, each image of the ear is modified with a different rotation angle. The accuracy of ear image recognition using GMI is 75.20% while using ZMI is 66%. The accuracy of using Geometric Moment Invariant is higher by 9.2% because Geometric Moment Invariant does not change the result of the image to be recognized due to translation, scaling, reflection and rotation.GMI cannot recognize the image of the ear with an average of 24.80% while the ZMI cannot recognize the image of the ear with an average of 34.40%. From the test results, at a rotation angle of 30 ° CCW, the geometric moment method has the best accuracy, while the Zernike moment method has the best accuracy at 30 ° CW. The best angle recognized by these two methods is 30°.

Prediction of the Number of Airport Passengers Using Fuzzy C-Means and Adaptive Neuro Fuzzy Inference System

by Suharjito MT and sunarsan sitohang

– Airport requires a system to predict the number of passengers as a reference for airport develo... more – Airport requires a system to predict the number of passengers as a reference for airport development planning. In this study, the data used are time series of the number of passengers for eleven years. These data will form patterns which indicate the number of passengers each month in a year as the input data and the number of passengers next year as a target prediction. After the input data are clustered into three types using fuzzy C-means (FCM), the data are processed using adaptive neuro fuzzy inference system (ANFIS) to get the prediction data. The result shows that the " Mean Absolute Percentage Errors (MAPE) which represent the errors for 4 years are 4.20%, 5.70%, 5.36% and 4.47% with an average of 4.93%. Based on this result, FCM and ANFIS can be combined to predict the data time series.

CONTENT BASED IMAGE RETRIEVAL USING BAG OF VISUAL WORDS AND MULTICLASS SUPPORT VECTOR MACHINE

Content Based Image Retrieval (CBIR) is an image retrieval method that is based on the meta conte... more Content Based Image Retrieval (CBIR) is an image retrieval method that is based on the meta content of an image which is proven to be more efficient than image retrieval method that is based on text or key words. Bag of Visual Words (BoVW) is a popular approach and the most widely used in CBIR problem. In this research, a CBIR method using BoVW and multiclass SVM classifier will be proposed. In BoVW, Scale Invariant Feature Transform (SIFT) will be used as the local features descriptor. This research will use Gaussian Mixture Model (GMM) as the method to do the visual vocabulary generation and Fisher Vector (FV) to create the encoder. The multiclass SVM classifier will use linear kernel, Hellinger's kernel, and chi-square kernel to do the classification. After the classification, the proposed CBIR will use the classification model to classify query image. After the query image class is known, then the color histogram features will be extracted from query image and dataset which only consists of image in the same class as query image. Datasets used in the research are Corel and Guang-Hai Liu (GHIM-10K). The experimental results show that BoVW with GMM and FV produces higher accuracy than other encoding methods in BoVW which produce a good retrieval result in the CBIR. 1. Introduction. The rapid development of digital technology has caused the amount of image data to increase at a rapid pace almost equal to the number of text data. However, searching data in image data is not as easy as searching data in text data. Searching in image (image retrieval) becomes one of the most commonly researched topics because of the need of an effective method to do searching in image data [1]. In general image retrieval is divided into 2 methods, which are text based and content based. Recent research shows that text-based method is not effective to represent an image into a small set of keywords which then shifts recent research topics to content-based method [2]. The idea of Content-Based Image Retrieval (CBIR) is searching an image using Meta data from the image. CBIR uses low level features in the searching method such as color, texture, or shape [3]. In general, the features which CBIR uses can be categorized into 2 key features: global feature descriptors and local feature descriptors. Global feature descriptors describe an image as a whole, while local feature descriptors describe an image by image patch (a small pixel). Global feature descriptors have an advantage in terms of its computational speed but the drawback is the low accuracy. Global feature descriptors tend to fail in identifying important visual characteristics from an image. On the other hand, local feature descriptors produce better accuracy than the global feature descriptors mainly because the image is represented by features calculated from the patch in the image. The drawback of the local features is that the size of feature space is large and will become very huge for large image database [4].

Mobile Decision Support System to Determine Toddler's Nutrition using Fuzzy Sugeno

Determination of nutritional status is closely related to the determination of dietary patterns s... more Determination of nutritional status is closely related to the determination of dietary patterns should be given to infants. Nutrition is very important role in mental, physical development, and human productivity. In this study, the system based on android is developed to determine the nutritional status of infants by using Fuzzy Sugeno. Indicator variables are age, height, circle head, and body weight according to the male or female. In this study, the results of measurements of nutritional status of children with Fuzzy Sugenoare tested by comparing the nutritional quality of the data Posyandu toddler by using anthropometric tables. The results of the evaluation measurement accuracy in this application are compared with the results of manual calculation based infant growth charts according to WHO standards. Therefore, these applications can be used to help the community in monitoring the nutritional status of children so that the growth of children is more appropriate in line with expectations.

Family Relationship Identification by Using Extract Feature of Gray Level Co-occurrence Matrix (GLCM) Based on Parents and Children Fingerprint

This study aims to find out the relations correspondence by using Gray Level Co-occurrence Matrix... more This study aims to find out the relations correspondence by using Gray Level Co-occurrence Matrix (GLCM) feature on parents and children finger print. The analysis is conducted by using the finger print of parents and family in one family There are 30 families used as sample with 3 finger print consists of mothers, fathers, and children finger print. Fingerprints data were taken by fingerprint digital persona u are u 4500 SDK. Data analysis is conducted by finding the correlation value between parents and children fingerprint by using correlation coefficient that gained from extract feature GLCM, both for similar family and different family. The study shows that the use of GLCM Extract Feature, normality data, and Correlation Coefficient could identify the correspondence relations between parents and children fingerprint on similar and different family. GLCM with four features (correlation, homogeneity, energy and contrast) are used to give good result. The four sides (0 o , 45 o , 90 o and 135 o) are used. It shows that side 0 o gives the higher accurate identification compared to other sides. 1. INTRODUCTION Information technology increases so fast in some various aspect scuh asgovernment, factory, school, hospital and others. Thus, the security system such as fingerprint, authentication, and verification becomes an interesting issue. Fingerprint is categorized as one of the best identification tools for its uniqueness [1]-[2]. Many researches about combination development [3] and the accurate of extract feature [4] are conducted [3]. Moreover, texture analysis is conducted to gain the score that will be used in finding the correlation score of study object. The other research that used wavelet and SVD method to identify the impact of genetics to the fingerprints. The research was conducted at 50 different fingerprints families. It found a high accuracy with 85%. The results showed that there were relationships between father and son = 0.3007 and the mother and children = 0.3448. While the relations value obtained on another family with father and son; mother and children = 0.2726 = 0.1747 [5]. In this research, fingerprint identification using extract feature is done to recognize the relations correspondence between parents and children. The result analysis consists of relations correspondence between the similar families with others. Extract feature method in this study uses the Gray Level Co-occurrence Matrix (GLCM) which combining the correlation analysis.

Query Optimization Using Fuzzy Logic in Integrated Database

Query optimization in integrated database can't be separated from data processing method. In orde... more Query optimization in integrated database can't be separated from data processing method. In order to have faster query response time, a method to optimize queries is required. One of many methods that can be used for query optimization is using fuzzy logic with Tsukamoto inference system. Value set on each variable is defined membership functions and Tsukamoto inference system used in determining these rules or the terms of query results, then apply it into query method or query line structure. The application of fuzzy logic inference systems with Tsukamoto can accelerate query response time, and will have more significant difference when the amount of selected data is greater. Such data must be analyzed to be used to support business decision making. Analysis for any type of decisions must also be done by considering different factors. For certain types of transactions made by specific customers, negotiation must be done with the grantor, which slows down the transaction process, to accelerate this process; a categorical list of consumers is made. With the list of categories of consumers, trading decisions can be directly carried out by the system. But the list of categories of consumers obtained from querying the current data provide considerable comparative results with a list of categories created by the grantor's decision. This means that the decision must still be made manually, because the results from the system cannot be used to make decisions. The significant difference between the results of the list of categories of consumers of the system and a list of categories of consumers of the decision maker occurs because the system makes decisions based only on the exact figure (firmly) from the factors used as reference, while decision maker (human) makes a decision based on estimates mainly on factors that are considered more important. Therefore, it needs to be optimized in the query that is used to obtain the list of categories of consumers. Use factors that affect query response time, among others are: missing indexes, inexact statistics, badly written queries, deadlocks, cursors, excessive fragmentation of indexes, and frequent recompilation of queries [1]. Optimization of query response time, especially on a large database with the number of joins that a lot can be done by using a genetic algorithm to obtain the best selection methods and the fitness function used in processing and the process of chromosomal mutations will shorten the response time [2]. In this study, query optimization will be carried out by using fuzzy logic. The use (of) query with fuzzy logic will accelerate query processing time because of the limitations use on the data that will be selected [3]. The use of fuzzy logic can not only improve the response time, but can also improve the selection results data obtained primarily related to the fuzzy logic so as to produce a system that is flexible and can handle general human demand [4]. The use of fuzzy logic will be a solution to problems in this study because fuzzy logic provides results that are can also use uncertain or it can be said to be approaching the human mind [5].

Hybrid Disk Drive Configuration on Database Server Virtualization

Solid State Drive (SSD) is a revolutionary new storage technology. Enterprise storage system usin... more Solid State Drive (SSD) is a revolutionary new storage technology. Enterprise storage system using full SSD is still very expensive, while hard disk drive (HDD) is still widely used. This study discusses hybrid configuration storage in virtualized server database with benchmark against four hybrid storage configuration for four databases, ORACLE, SQL Server, MySQL and PostgreSQL on Windows Server virtualization. Benchmark using TPC-C and TPC-H to get the best performance of four configurations were tested. The results of this study indicate HDD storage configurations as visual disk drive OS and SSD as visual disk drives database get better performance as on line transaction processing (OLTP) and on line analytical processing (OLAP) database server compared with SSD as visual disk drive OS and HDD as a visual disk drive database. Based on the data research TPC-C, OLTP get best performance at HDD storage configurations as visual disk drive OS and SSD as a visual disk drives database and temporary files.

OLTP PERFORMANCE IMPROVEMENT USING FILE-SYSTEMS LAYER COMPRESSION

Graphical abstract Abstract Database for Online Transaction Processing (OLTP) application is used... more Graphical abstract Abstract Database for Online Transaction Processing (OLTP) application is used by almost every corporations that has adopted computerisation to support their operational day to day business. Compression in the storage or file-systems layer has not been widely adopted for OLTP database because of the concern that it might decrease database performance. OLTP compression in the database layer is available commercially but it has a significant licence cost that reduces the cost saving of compression. In this research, transparent file-system compression with LZ4, LZJB and ZLE algorithm have been tested to improve performance of OLTP application. Using Swing-bench as the benchmark tool and Oracle database 12c, The result indicated that on OLTP workload, LZJB was the most optimal compression algorithm with performance improvement up to 49% and consistent reduction of maximum response time and CPU utilisation overhead, while LZ4 was the compression with the highest compression ratio and ZLE was the compression with the lowest CPU utilisation overhead. In terms of compression ratio, LZ4 can deliver the highest compression ratio which is 5.32, followed by LZJB, 4.92; and ZLE, 1.76. Furthermore, it is found that there is indeed a risk of reduced performance and/or an increase of maximum response time.

Fuzzy Expert System for Classifying Pests and Diseases of Paddy Using Bee Colony Algorithms

by Suharjito MT and Yovita Tunardi

– Agriculture sector has an important role in Indonesia. However,there are still many problems to... more – Agriculture sector has an important role in Indonesia. However,there are still many problems to deal with, such as pests and diseases, especially for paddy which is an important crop because it is the main food for people in Indonesia. Many things have been done by the government to solve that problem, but it has not been resolved because of the lack of knowledge of Indonesian farmers. This research proposes a new way to solve that problem by applying optimized Fuzzy Logic (Fuzzy K-Means and Fuzzy C-Means) with Bee Colony to get pest and disease classifications for paddy based on existing symptoms. Bee Colony will be used to determine the initial centroid for Fuzzy Logic. The proposed approach is evaluated based on the accuracy by comparing the value of Mean Square Error (MSE) and the accuracy of every method for pest and disease classifications. MSE of the new model is smaller than regular Fuzzy Logic by 0.016 for FCM and 1.213 for FKM. The accuracy of the new model is greater than regular Fuzzy Logic by 13.87% for FCM and 18.39% for FKM. It can be concluded that Bee Colony can be used to optimize Fuzzy Logic for pest and disease classifications for paddy with better classification results.

Hand Motion Gesture for Human-Computer Interaction Using Support Vector Machine and Hidden Markov Model

– Hand gesture recognition for human-computer interaction has become very popular in recent times... more – Hand gesture recognition for human-computer interaction has become very popular in recent times. The main problem of this technology is how the system can recognize the presence of a gesture in a streaming video. In this paper, we propose a model that can recognize hand motion gesture in avideo stream using Support Vector Machine and Hidden-Markov model. Support Vector Machine has the advantage of generalizing classification. On the other hand, Hidden Markov Model is a statistical model that is capable of modeling Spatial-temporal time series. This system is divided into two main processes. First, this system recognizes hand posture using SMV (static gesture recognition) and generates sequence observation which is used for the second process later. The second process is recognizing dynamic hand gesture with the sequence observation from the static gesture. The implementation shows that static hand gesture recognition achieves average accuracy at 91% using testing dataset. Meanwhile,dynamically isolated hand gesture recognition gets average accuracy at 89% using testing dataset. We also have tested the system with continuous dynamic gesture using video stream. The system can recognize the gesture very well with accuracy of 83%. This achievement shows that the model can be used in human computer-interaction with specific supports such as Vector Machine and Hidden Markov Model parameter.

File Synchronization Algorithm Using Hash Transfer Synchronization Method in Distributed System

by Suharjito MT, Suharjito Suharjito, and Aryo Nugroho

– One of the problems in a distributed system is how to synchronize files in the system that are ... more – One of the problems in a distributed system is how to synchronize files in the system that are located separately. Many ways can be used to detect files that are unsynchronized with file references of the source server. This article proposes a method that can detect and fix unsynchronized files quickly although the network has low-performance traffic using Hash Transfer Synchronization algorithm. The result of this study is a system that can be used to detect unsynchronized files and fix approximately 14 million files at 1400 servers within 8 hours in the distributed system.

References of Book-Teknik dan Analisis Pengambilan Keputusan Fuzzy Dalam Manajemen Rantai Pasok

by Retno Astuti and Suharjito MT

Teknik dan Analisis Pengambilan Keputusan Fuzzy dalam Manajemen Rantai Pasok Sistem Manajemen Ran... more Teknik dan Analisis Pengambilan Keputusan Fuzzy dalam Manajemen Rantai Pasok Sistem Manajemen Rantai Pasok merupakan pendekatan untuk mengintegrasikan pemasok, pengusaha, sistem produksi, distribusi, logistik dan pemasaran kedalam sistem manajemen perusahaan secara efektif dan efisien. Rangkaian aktivitas-aktivitas rantai pasok melibatkan banyak hal yang memerlukan pengambilan keputusan sederhana sampai kompleks yang ranah substansinya sering tidak pasti dan kabur (fuzzy) sehingga sulit direpresentasikan dan disimpulkan. Buku ini membahas teknik dan analisis pengambilan keputusan fuzzy dalam manajemen rantai pasok. Bahasan mencakup konsep, teori dan sistem keputusan, sistem manajemen rantai pasok pertanian, konsep dan struktur sistem fuzzy, pengembangan teknik penelitian operasional: pemrograman linier fuzzy, analisis risiko fuzzy, analisis kelayakan investasi fuzzy, proses hirarki analitik (AHP) fuzzy dan regresi fuzzy. Teknik-teknik tersebut dapat dijadikan model dasar pengembanga...

DSS for Agricultural Products Supply Chain Risk Balancing using Stakeholder Dialogues and Fuzzy Non Linear Regression

by Suharjito MT, Marimin Marimin, and Suharjito Suharjito

The high complexity of the supply chain network and the characteristics of products made supply c... more The high complexity of the supply chain network and the characteristics of products made supply chain management of agricultural products was more susceptible to the risks emergence of loss. Therefore, it is required to develop a mechanism for price negotiation which distributes the risks fairly for each stakeholder in the supply chain. In addition it is necessary to identify and evaluate supply chain risks in order to avoid continuing problems that can occur at any point in the supply chain network. The objectives of this study were to describe the model of identification and evaluation for corn supply chain risk, to formulate a fair pricing mechanism for corn supply chain using risk balancing model. Risk identification was conducted using fuzzy Analytical Hierarchy Process (AHP) approach and risk evaluation was done by using fuzzy logic with data input form the opinion of several experts maize supply chain. A fairly pricing model at farmer level was developed by using stakeholder dialogue approach based on a balanced fuzzy risk utility preference that was faced by all stages of the supply chain. In addition, fuzzy risk utility optimization was used to get a consensus of the supply chain stakeholder dialogue, where basic risk utility function was derived using fuzzy regression approach. Risk mitigation for each stage of supply chain was developed using fuzzy inferences based on the risk that has been evaluated. Based on the verification results, the model could identify the level of risks for each party of the supply chain and the action that must be taken for minimizing its impacts using appropriate strategies. The model could shift the risks from the farmer to the other parties to determine the fair benefit distribution on the price negotiation.
Keywords: supply chain, risk balancing, fuzzy utility, optimization, stakeholder dialogue

Simulation Model of Bus Rapid Transit

by Suharjito MT and FERGYANTO GUNAWAN

Bus rapid transit system is modern solution for mass transportation system. The system, in compar... more Bus rapid transit system is modern solution for mass transportation system. The system, in
comparison to the rail-based transportation system, is significantly cheaper and requires shorter development
time, but lower performance. The BRT system performance strongly depends on variables related to station
design and infrastructure. A numerical model offers an effective and efficient means to evaluate the system
performance. This article offers a detailed numerical model on the basis of the discrete-event approach and
demonstrates its application.

Intelligent Decision Support System Modeling for Agri-Supply Chain Risk Balancing on Price Determination

by Suharjito MT and machfud machfud

Generally, it is found that farmers have no bargaining power in price determination. Thus, they h... more Generally, it is found that farmers have no bargaining power in price determination. Thus, they have to bear all the risks compared to the others. In addition, price exchanging at farmer level tends to fluctuate significantly. Therefore, it is required to develop a mechanism for price negotiation, which distributes the risks fairly for each stakeholder in the supply chain. The objective of this study is to describe the model of identification and evaluation for corn supply chain risks, to formulate a fair pricing mechanism for corn supply chain using risk balancing model, and to develop intelligent decision support system for supply chain risk management. Risk identification was conducted using fuzzy AHP approach and risk evaluation was done by using fuzzy logic with data input derived from the opinion of several experts of maize supply chain. A fairly pricing model at farmer level was developed by using stakeholder dialogue approach based on a balanced fuzzy risk utility preference that faced by all stages of the supply chain. In addition, fuzzy risk utility optimization was used to get a consensus of the supply chain stakeholder dialogue, where basic risk utility function was derived using fuzzy regression approach. Risk mitigation for each stage of supply chain was developed using fuzzy inferences based on the risk that has been evaluated. Based on the verification results, the model could identify the level of risks for each party of the supply chain and the action that must be taken for minimizing its impacts using appropriate strategies. The model can shift the risks from the farmer to the other parties to determine the fair benefit distribution on the price negotiation

DYNAMIC DIFFICULTY ADJUSTMENT IN GAME BASED ON TYPE OF PLAYER WITH ANFIS METHOD

by Suharjito MT and Kurniawan Sutanto

The purpose of this research is to measure the accuracy and effectiveness of Adaptive Neuro Fuzzy... more The purpose of this research is to measure the accuracy and effectiveness of Adaptive Neuro Fuzzy Inference System (ANFIS) when it implemented as Dynamic Difficulty Adjustment, to manage the difficulty of a game according to the player's behavior, as the main problem is players need to switch the difficulty level manually, and it will be annoyed them. The inputs are the amount of enemies defeated and hero taken damages. There are five levels of difficulty. The measurement of accuracy done by Mean-Squared Error, meanwhile the measurement of effectiveness done by comparing the value of Mean-Squared Error to the Artificial Neural Network's. The result is ANFIS has better performance on accuracy and effectiveness, as the value of MSE is 0.010592338 compared to 0.026847068 of ANN, so it can be conclude that ANFIS has good accuracy and effectiveness when it implemented as Dynamic Difficulty Adjustment.

PLAGIARISM DETECTION ALGORITHM USING NATURAL LANGUAGE PROCESSING BASED ON GRAMMAR ANALYZING

by Suharjito MT and Angry Ronald

Plagiarism has become one of the most concerned problems since there are several kinds of plagiar... more Plagiarism has become one of the most concerned problems since there are several kinds of plagiarism that are hard to detect. Extrinsic plagiarism is now being handled well, but intrinsic plagiarism is not. Intrinsic plagiarism detection is being distracted by the mixed up structure and the using of another word which have the same meaning. Several methods have been research to handle this problem, not only using the pattern reading but also parsing the sentences, but still couldn't give one more accurate way to detect it. In this research, we propose to use Natural Language Processing (NLP) to create the new way to detect plagiarism. Begin with using syntactic parsing method to parse the suspicious document and find the list of words which have the same meaning with it while considering the Part-Of-Speech (POS) element of that word (semantic parsing). This algorithm also includes creating the new structure of the object before comparing them. The result of this research presents the accuracy comparison between Ferret, WCopyFind, and this algorithm. This algorithm gives the significant way to detect the plagiarism which is proven by T-Test.

Choosing Mutual Fund based on Purpose, Risk Level and Inventor's Profile using Neutral Network

by Bambang H. Iswanto, Suharjito MT, and Suharjito Suharjito

Investment in financial asset these days are easily to be done. The principle in investment is hi... more Investment in financial asset these days are easily to be done. The principle in investment is higher return, higher risk. An investment with the very high return, contain very high level of risk. Otherwise if you invest in the financial asset that contain low risk level, then the expected return will be low. Therefore decision of choosing the right investment instrument is very important because it is related to risk, individual readiness in its implementation and suitability of investor profile itself. This thesis propose a machine learning method to develop a decission support system for investment manager in determining the suitable investment instrument for the individual client based on financial objective , risk level and investment period. Machine learning was choosen because it was known of its ability in recognizing the complex pattern based on learning process using the given data set. Artificial neural network multilayer perceptron (MLP) with two layers could be used as a classificator model to predict the mutual fund investment type that was suitable for investor based on investment purpose , risk level and investor profile. Experiment using 2-layers artificial neural network, 9 unit input, 16 hidden units and 4 output units that was trained with 50 data points consists of 9 vestor input component and 4 vector target components. The architecture and the data in the network could done the classification with accuration 93.33 percents.

PREDICTION OF RESEARCH TOPICS USING COMBINATION OF MACHINE LEARNING AND LOGISTIC CURVE

Prediction of future research topics is usually undertaken, especially by policy makers in resear... more Prediction of future research topics is usually undertaken, especially by policy makers in research environment, to determine the focus of future studies. Analysis of future research topics by using a numerical approach based on scientific publications or patents have been previously done by several researchers mostly by classifying the topics based on the similarity of the documents. Meanwhile, time series analysis using statistical techniques, machine learning and logistic curves has also been conducted but for areas other than research topics. In this study, we will perform time series analysis to predict the topic of research by employing combination of machine learning approaches such as Neural Network, Extreme Learning Machine and Support Vector Machine. The prediction result is then finally refined by logistic curve. The dataset used in this study is a research report on Bioinformatics from Microsoft Research and NCBI (National Center for Biotechnology Information), over the past 30 years. Experimental result indicates that the combination of machine learning approaches and logistic-curve may improve the prediction accuracy. In addition, the emerging topic of the same dataset can be predicted with a relatively high average precision. Based on the dataset used, some emerging topics predicted for the next three years are 'Sequence Alignment', 'Index Terms', 'Secondary Structure', 'Nucleic Acid', and 'Saccharomyces cerevisiae'

Adaptive Neuro-Fuzzy Inference System for Texture Image Classification

—One of the most important problems in pattern recognition is texture-based image classification.... more —One of the most important problems in pattern recognition is texture-based image classification. In this paper , the combination of Discrete Cosine Transform (DCT) and Gray Level Co-Occurrence Matrix (GLCM) methods for feature extraction was proposed. The attributes extracted from DCT method were mean and variance, while the attributes extracted from GLCM method were energy and entropy. Adaptive Neuro-Fuzzy Inference System (ANFIS) was used as a classifier. The classifier model was trained using 50% of the texture images and remaining images were used for testing. Four classes of texture images were downloaded from KTH-TIPS (Textures under varying Illumination, Pose and Scale) image database, three of which were used in each experiments thus there were four data combination. The best data testing accuracy result towards textures of crumpled aluminium foil, corduroy, and orange peel is 98.3%, which is 1.6% better than one hidden-layer feed forward neural network classifier. In average, testing accuracy result of ANFIS excelled one hidden-layer feed forward neural network with 93.7% over 90.4%.

Human Ear Recognition Methods Based on Image Rotation

– The ear is part of biometrics that has a unique and stable structure, avoided from aging. This ... more – The ear is part of biometrics that has a unique and stable structure, avoided from aging. This study aims to look at the accuracy of Geometric Moment invariant (GMI) and Zernike Moment invariant (ZMI) with Self-Organizing Maps to recognize human ears. This study used data from ear AMI database, consisting of 125 ears from 25 different people, in which each person has 5 images of the right ear. In this study, each image of the ear is modified with a different rotation angle. The accuracy of ear image recognition using GMI is 75.20% while using ZMI is 66%. The accuracy of using Geometric Moment Invariant is higher by 9.2% because Geometric Moment Invariant does not change the result of the image to be recognized due to translation, scaling, reflection and rotation.GMI cannot recognize the image of the ear with an average of 24.80% while the ZMI cannot recognize the image of the ear with an average of 34.40%. From the test results, at a rotation angle of 30 ° CCW, the geometric moment method has the best accuracy, while the Zernike moment method has the best accuracy at 30 ° CW. The best angle recognized by these two methods is 30°.

Prediction of the Number of Airport Passengers Using Fuzzy C-Means and Adaptive Neuro Fuzzy Inference System

by Suharjito MT and sunarsan sitohang

– Airport requires a system to predict the number of passengers as a reference for airport develo... more – Airport requires a system to predict the number of passengers as a reference for airport development planning. In this study, the data used are time series of the number of passengers for eleven years. These data will form patterns which indicate the number of passengers each month in a year as the input data and the number of passengers next year as a target prediction. After the input data are clustered into three types using fuzzy C-means (FCM), the data are processed using adaptive neuro fuzzy inference system (ANFIS) to get the prediction data. The result shows that the " Mean Absolute Percentage Errors (MAPE) which represent the errors for 4 years are 4.20%, 5.70%, 5.36% and 4.47% with an average of 4.93%. Based on this result, FCM and ANFIS can be combined to predict the data time series.

CONTENT BASED IMAGE RETRIEVAL USING BAG OF VISUAL WORDS AND MULTICLASS SUPPORT VECTOR MACHINE

Content Based Image Retrieval (CBIR) is an image retrieval method that is based on the meta conte... more Content Based Image Retrieval (CBIR) is an image retrieval method that is based on the meta content of an image which is proven to be more efficient than image retrieval method that is based on text or key words. Bag of Visual Words (BoVW) is a popular approach and the most widely used in CBIR problem. In this research, a CBIR method using BoVW and multiclass SVM classifier will be proposed. In BoVW, Scale Invariant Feature Transform (SIFT) will be used as the local features descriptor. This research will use Gaussian Mixture Model (GMM) as the method to do the visual vocabulary generation and Fisher Vector (FV) to create the encoder. The multiclass SVM classifier will use linear kernel, Hellinger's kernel, and chi-square kernel to do the classification. After the classification, the proposed CBIR will use the classification model to classify query image. After the query image class is known, then the color histogram features will be extracted from query image and dataset which only consists of image in the same class as query image. Datasets used in the research are Corel and Guang-Hai Liu (GHIM-10K). The experimental results show that BoVW with GMM and FV produces higher accuracy than other encoding methods in BoVW which produce a good retrieval result in the CBIR. 1. Introduction. The rapid development of digital technology has caused the amount of image data to increase at a rapid pace almost equal to the number of text data. However, searching data in image data is not as easy as searching data in text data. Searching in image (image retrieval) becomes one of the most commonly researched topics because of the need of an effective method to do searching in image data [1]. In general image retrieval is divided into 2 methods, which are text based and content based. Recent research shows that text-based method is not effective to represent an image into a small set of keywords which then shifts recent research topics to content-based method [2]. The idea of Content-Based Image Retrieval (CBIR) is searching an image using Meta data from the image. CBIR uses low level features in the searching method such as color, texture, or shape [3]. In general, the features which CBIR uses can be categorized into 2 key features: global feature descriptors and local feature descriptors. Global feature descriptors describe an image as a whole, while local feature descriptors describe an image by image patch (a small pixel). Global feature descriptors have an advantage in terms of its computational speed but the drawback is the low accuracy. Global feature descriptors tend to fail in identifying important visual characteristics from an image. On the other hand, local feature descriptors produce better accuracy than the global feature descriptors mainly because the image is represented by features calculated from the patch in the image. The drawback of the local features is that the size of feature space is large and will become very huge for large image database [4].

Mobile Decision Support System to Determine Toddler's Nutrition using Fuzzy Sugeno

Determination of nutritional status is closely related to the determination of dietary patterns s... more Determination of nutritional status is closely related to the determination of dietary patterns should be given to infants. Nutrition is very important role in mental, physical development, and human productivity. In this study, the system based on android is developed to determine the nutritional status of infants by using Fuzzy Sugeno. Indicator variables are age, height, circle head, and body weight according to the male or female. In this study, the results of measurements of nutritional status of children with Fuzzy Sugenoare tested by comparing the nutritional quality of the data Posyandu toddler by using anthropometric tables. The results of the evaluation measurement accuracy in this application are compared with the results of manual calculation based infant growth charts according to WHO standards. Therefore, these applications can be used to help the community in monitoring the nutritional status of children so that the growth of children is more appropriate in line with expectations.

Family Relationship Identification by Using Extract Feature of Gray Level Co-occurrence Matrix (GLCM) Based on Parents and Children Fingerprint

This study aims to find out the relations correspondence by using Gray Level Co-occurrence Matrix... more This study aims to find out the relations correspondence by using Gray Level Co-occurrence Matrix (GLCM) feature on parents and children finger print. The analysis is conducted by using the finger print of parents and family in one family There are 30 families used as sample with 3 finger print consists of mothers, fathers, and children finger print. Fingerprints data were taken by fingerprint digital persona u are u 4500 SDK. Data analysis is conducted by finding the correlation value between parents and children fingerprint by using correlation coefficient that gained from extract feature GLCM, both for similar family and different family. The study shows that the use of GLCM Extract Feature, normality data, and Correlation Coefficient could identify the correspondence relations between parents and children fingerprint on similar and different family. GLCM with four features (correlation, homogeneity, energy and contrast) are used to give good result. The four sides (0 o , 45 o , 90 o and 135 o) are used. It shows that side 0 o gives the higher accurate identification compared to other sides. 1. INTRODUCTION Information technology increases so fast in some various aspect scuh asgovernment, factory, school, hospital and others. Thus, the security system such as fingerprint, authentication, and verification becomes an interesting issue. Fingerprint is categorized as one of the best identification tools for its uniqueness [1]-[2]. Many researches about combination development [3] and the accurate of extract feature [4] are conducted [3]. Moreover, texture analysis is conducted to gain the score that will be used in finding the correlation score of study object. The other research that used wavelet and SVD method to identify the impact of genetics to the fingerprints. The research was conducted at 50 different fingerprints families. It found a high accuracy with 85%. The results showed that there were relationships between father and son = 0.3007 and the mother and children = 0.3448. While the relations value obtained on another family with father and son; mother and children = 0.2726 = 0.1747 [5]. In this research, fingerprint identification using extract feature is done to recognize the relations correspondence between parents and children. The result analysis consists of relations correspondence between the similar families with others. Extract feature method in this study uses the Gray Level Co-occurrence Matrix (GLCM) which combining the correlation analysis.

Query Optimization Using Fuzzy Logic in Integrated Database

Query optimization in integrated database can't be separated from data processing method. In orde... more Query optimization in integrated database can't be separated from data processing method. In order to have faster query response time, a method to optimize queries is required. One of many methods that can be used for query optimization is using fuzzy logic with Tsukamoto inference system. Value set on each variable is defined membership functions and Tsukamoto inference system used in determining these rules or the terms of query results, then apply it into query method or query line structure. The application of fuzzy logic inference systems with Tsukamoto can accelerate query response time, and will have more significant difference when the amount of selected data is greater. Such data must be analyzed to be used to support business decision making. Analysis for any type of decisions must also be done by considering different factors. For certain types of transactions made by specific customers, negotiation must be done with the grantor, which slows down the transaction process, to accelerate this process; a categorical list of consumers is made. With the list of categories of consumers, trading decisions can be directly carried out by the system. But the list of categories of consumers obtained from querying the current data provide considerable comparative results with a list of categories created by the grantor's decision. This means that the decision must still be made manually, because the results from the system cannot be used to make decisions. The significant difference between the results of the list of categories of consumers of the system and a list of categories of consumers of the decision maker occurs because the system makes decisions based only on the exact figure (firmly) from the factors used as reference, while decision maker (human) makes a decision based on estimates mainly on factors that are considered more important. Therefore, it needs to be optimized in the query that is used to obtain the list of categories of consumers. Use factors that affect query response time, among others are: missing indexes, inexact statistics, badly written queries, deadlocks, cursors, excessive fragmentation of indexes, and frequent recompilation of queries [1]. Optimization of query response time, especially on a large database with the number of joins that a lot can be done by using a genetic algorithm to obtain the best selection methods and the fitness function used in processing and the process of chromosomal mutations will shorten the response time [2]. In this study, query optimization will be carried out by using fuzzy logic. The use (of) query with fuzzy logic will accelerate query processing time because of the limitations use on the data that will be selected [3]. The use of fuzzy logic can not only improve the response time, but can also improve the selection results data obtained primarily related to the fuzzy logic so as to produce a system that is flexible and can handle general human demand [4]. The use of fuzzy logic will be a solution to problems in this study because fuzzy logic provides results that are can also use uncertain or it can be said to be approaching the human mind [5].

Hybrid Disk Drive Configuration on Database Server Virtualization

Solid State Drive (SSD) is a revolutionary new storage technology. Enterprise storage system usin... more Solid State Drive (SSD) is a revolutionary new storage technology. Enterprise storage system using full SSD is still very expensive, while hard disk drive (HDD) is still widely used. This study discusses hybrid configuration storage in virtualized server database with benchmark against four hybrid storage configuration for four databases, ORACLE, SQL Server, MySQL and PostgreSQL on Windows Server virtualization. Benchmark using TPC-C and TPC-H to get the best performance of four configurations were tested. The results of this study indicate HDD storage configurations as visual disk drive OS and SSD as visual disk drives database get better performance as on line transaction processing (OLTP) and on line analytical processing (OLAP) database server compared with SSD as visual disk drive OS and HDD as a visual disk drive database. Based on the data research TPC-C, OLTP get best performance at HDD storage configurations as visual disk drive OS and SSD as a visual disk drives database and temporary files.

OLTP PERFORMANCE IMPROVEMENT USING FILE-SYSTEMS LAYER COMPRESSION

Graphical abstract Abstract Database for Online Transaction Processing (OLTP) application is used... more Graphical abstract Abstract Database for Online Transaction Processing (OLTP) application is used by almost every corporations that has adopted computerisation to support their operational day to day business. Compression in the storage or file-systems layer has not been widely adopted for OLTP database because of the concern that it might decrease database performance. OLTP compression in the database layer is available commercially but it has a significant licence cost that reduces the cost saving of compression. In this research, transparent file-system compression with LZ4, LZJB and ZLE algorithm have been tested to improve performance of OLTP application. Using Swing-bench as the benchmark tool and Oracle database 12c, The result indicated that on OLTP workload, LZJB was the most optimal compression algorithm with performance improvement up to 49% and consistent reduction of maximum response time and CPU utilisation overhead, while LZ4 was the compression with the highest compression ratio and ZLE was the compression with the lowest CPU utilisation overhead. In terms of compression ratio, LZ4 can deliver the highest compression ratio which is 5.32, followed by LZJB, 4.92; and ZLE, 1.76. Furthermore, it is found that there is indeed a risk of reduced performance and/or an increase of maximum response time.

Fuzzy Expert System for Classifying Pests and Diseases of Paddy Using Bee Colony Algorithms

by Suharjito MT and Yovita Tunardi

– Agriculture sector has an important role in Indonesia. However,there are still many problems to... more – Agriculture sector has an important role in Indonesia. However,there are still many problems to deal with, such as pests and diseases, especially for paddy which is an important crop because it is the main food for people in Indonesia. Many things have been done by the government to solve that problem, but it has not been resolved because of the lack of knowledge of Indonesian farmers. This research proposes a new way to solve that problem by applying optimized Fuzzy Logic (Fuzzy K-Means and Fuzzy C-Means) with Bee Colony to get pest and disease classifications for paddy based on existing symptoms. Bee Colony will be used to determine the initial centroid for Fuzzy Logic. The proposed approach is evaluated based on the accuracy by comparing the value of Mean Square Error (MSE) and the accuracy of every method for pest and disease classifications. MSE of the new model is smaller than regular Fuzzy Logic by 0.016 for FCM and 1.213 for FKM. The accuracy of the new model is greater than regular Fuzzy Logic by 13.87% for FCM and 18.39% for FKM. It can be concluded that Bee Colony can be used to optimize Fuzzy Logic for pest and disease classifications for paddy with better classification results.

Hand Motion Gesture for Human-Computer Interaction Using Support Vector Machine and Hidden Markov Model

– Hand gesture recognition for human-computer interaction has become very popular in recent times... more – Hand gesture recognition for human-computer interaction has become very popular in recent times. The main problem of this technology is how the system can recognize the presence of a gesture in a streaming video. In this paper, we propose a model that can recognize hand motion gesture in avideo stream using Support Vector Machine and Hidden-Markov model. Support Vector Machine has the advantage of generalizing classification. On the other hand, Hidden Markov Model is a statistical model that is capable of modeling Spatial-temporal time series. This system is divided into two main processes. First, this system recognizes hand posture using SMV (static gesture recognition) and generates sequence observation which is used for the second process later. The second process is recognizing dynamic hand gesture with the sequence observation from the static gesture. The implementation shows that static hand gesture recognition achieves average accuracy at 91% using testing dataset. Meanwhile,dynamically isolated hand gesture recognition gets average accuracy at 89% using testing dataset. We also have tested the system with continuous dynamic gesture using video stream. The system can recognize the gesture very well with accuracy of 83%. This achievement shows that the model can be used in human computer-interaction with specific supports such as Vector Machine and Hidden Markov Model parameter.

File Synchronization Algorithm Using Hash Transfer Synchronization Method in Distributed System

by Suharjito MT, Suharjito Suharjito, and Aryo Nugroho

– One of the problems in a distributed system is how to synchronize files in the system that are ... more – One of the problems in a distributed system is how to synchronize files in the system that are located separately. Many ways can be used to detect files that are unsynchronized with file references of the source server. This article proposes a method that can detect and fix unsynchronized files quickly although the network has low-performance traffic using Hash Transfer Synchronization algorithm. The result of this study is a system that can be used to detect unsynchronized files and fix approximately 14 million files at 1400 servers within 8 hours in the distributed system.

References of Book-Teknik dan Analisis Pengambilan Keputusan Fuzzy Dalam Manajemen Rantai Pasok

by Retno Astuti and Suharjito MT

Teknik dan Analisis Pengambilan Keputusan Fuzzy dalam Manajemen Rantai Pasok Sistem Manajemen Ran... more Teknik dan Analisis Pengambilan Keputusan Fuzzy dalam Manajemen Rantai Pasok Sistem Manajemen Rantai Pasok merupakan pendekatan untuk mengintegrasikan pemasok, pengusaha, sistem produksi, distribusi, logistik dan pemasaran kedalam sistem manajemen perusahaan secara efektif dan efisien. Rangkaian aktivitas-aktivitas rantai pasok melibatkan banyak hal yang memerlukan pengambilan keputusan sederhana sampai kompleks yang ranah substansinya sering tidak pasti dan kabur (fuzzy) sehingga sulit direpresentasikan dan disimpulkan. Buku ini membahas teknik dan analisis pengambilan keputusan fuzzy dalam manajemen rantai pasok. Bahasan mencakup konsep, teori dan sistem keputusan, sistem manajemen rantai pasok pertanian, konsep dan struktur sistem fuzzy, pengembangan teknik penelitian operasional: pemrograman linier fuzzy, analisis risiko fuzzy, analisis kelayakan investasi fuzzy, proses hirarki analitik (AHP) fuzzy dan regresi fuzzy. Teknik-teknik tersebut dapat dijadikan model dasar pengembanga...

DSS for Agricultural Products Supply Chain Risk Balancing using Stakeholder Dialogues and Fuzzy Non Linear Regression

by Suharjito MT, Marimin Marimin, and Suharjito Suharjito

The high complexity of the supply chain network and the characteristics of products made supply c... more The high complexity of the supply chain network and the characteristics of products made supply chain management of agricultural products was more susceptible to the risks emergence of loss. Therefore, it is required to develop a mechanism for price negotiation which distributes the risks fairly for each stakeholder in the supply chain. In addition it is necessary to identify and evaluate supply chain risks in order to avoid continuing problems that can occur at any point in the supply chain network. The objectives of this study were to describe the model of identification and evaluation for corn supply chain risk, to formulate a fair pricing mechanism for corn supply chain using risk balancing model. Risk identification was conducted using fuzzy Analytical Hierarchy Process (AHP) approach and risk evaluation was done by using fuzzy logic with data input form the opinion of several experts maize supply chain. A fairly pricing model at farmer level was developed by using stakeholder dialogue approach based on a balanced fuzzy risk utility preference that was faced by all stages of the supply chain. In addition, fuzzy risk utility optimization was used to get a consensus of the supply chain stakeholder dialogue, where basic risk utility function was derived using fuzzy regression approach. Risk mitigation for each stage of supply chain was developed using fuzzy inferences based on the risk that has been evaluated. Based on the verification results, the model could identify the level of risks for each party of the supply chain and the action that must be taken for minimizing its impacts using appropriate strategies. The model could shift the risks from the farmer to the other parties to determine the fair benefit distribution on the price negotiation.
Keywords: supply chain, risk balancing, fuzzy utility, optimization, stakeholder dialogue

Simulation Model of Bus Rapid Transit

by Suharjito MT and FERGYANTO GUNAWAN

Bus rapid transit system is modern solution for mass transportation system. The system, in compar... more Bus rapid transit system is modern solution for mass transportation system. The system, in
comparison to the rail-based transportation system, is significantly cheaper and requires shorter development
time, but lower performance. The BRT system performance strongly depends on variables related to station
design and infrastructure. A numerical model offers an effective and efficient means to evaluate the system
performance. This article offers a detailed numerical model on the basis of the discrete-event approach and
demonstrates its application.

Intelligent Decision Support System Modeling for Agri-Supply Chain Risk Balancing on Price Determination

by Suharjito MT and machfud machfud

Generally, it is found that farmers have no bargaining power in price determination. Thus, they h... more Generally, it is found that farmers have no bargaining power in price determination. Thus, they have to bear all the risks compared to the others. In addition, price exchanging at farmer level tends to fluctuate significantly. Therefore, it is required to develop a mechanism for price negotiation, which distributes the risks fairly for each stakeholder in the supply chain. The objective of this study is to describe the model of identification and evaluation for corn supply chain risks, to formulate a fair pricing mechanism for corn supply chain using risk balancing model, and to develop intelligent decision support system for supply chain risk management. Risk identification was conducted using fuzzy AHP approach and risk evaluation was done by using fuzzy logic with data input derived from the opinion of several experts of maize supply chain. A fairly pricing model at farmer level was developed by using stakeholder dialogue approach based on a balanced fuzzy risk utility preference that faced by all stages of the supply chain. In addition, fuzzy risk utility optimization was used to get a consensus of the supply chain stakeholder dialogue, where basic risk utility function was derived using fuzzy regression approach. Risk mitigation for each stage of supply chain was developed using fuzzy inferences based on the risk that has been evaluated. Based on the verification results, the model could identify the level of risks for each party of the supply chain and the action that must be taken for minimizing its impacts using appropriate strategies. The model can shift the risks from the farmer to the other parties to determine the fair benefit distribution on the price negotiation

DYNAMIC DIFFICULTY ADJUSTMENT IN GAME BASED ON TYPE OF PLAYER WITH ANFIS METHOD

by Suharjito MT and Kurniawan Sutanto

The purpose of this research is to measure the accuracy and effectiveness of Adaptive Neuro Fuzzy... more The purpose of this research is to measure the accuracy and effectiveness of Adaptive Neuro Fuzzy Inference System (ANFIS) when it implemented as Dynamic Difficulty Adjustment, to manage the difficulty of a game according to the player's behavior, as the main problem is players need to switch the difficulty level manually, and it will be annoyed them. The inputs are the amount of enemies defeated and hero taken damages. There are five levels of difficulty. The measurement of accuracy done by Mean-Squared Error, meanwhile the measurement of effectiveness done by comparing the value of Mean-Squared Error to the Artificial Neural Network's. The result is ANFIS has better performance on accuracy and effectiveness, as the value of MSE is 0.010592338 compared to 0.026847068 of ANN, so it can be conclude that ANFIS has good accuracy and effectiveness when it implemented as Dynamic Difficulty Adjustment.

PLAGIARISM DETECTION ALGORITHM USING NATURAL LANGUAGE PROCESSING BASED ON GRAMMAR ANALYZING

by Suharjito MT and Angry Ronald

Plagiarism has become one of the most concerned problems since there are several kinds of plagiar... more Plagiarism has become one of the most concerned problems since there are several kinds of plagiarism that are hard to detect. Extrinsic plagiarism is now being handled well, but intrinsic plagiarism is not. Intrinsic plagiarism detection is being distracted by the mixed up structure and the using of another word which have the same meaning. Several methods have been research to handle this problem, not only using the pattern reading but also parsing the sentences, but still couldn't give one more accurate way to detect it. In this research, we propose to use Natural Language Processing (NLP) to create the new way to detect plagiarism. Begin with using syntactic parsing method to parse the suspicious document and find the list of words which have the same meaning with it while considering the Part-Of-Speech (POS) element of that word (semantic parsing). This algorithm also includes creating the new structure of the object before comparing them. The result of this research presents the accuracy comparison between Ferret, WCopyFind, and this algorithm. This algorithm gives the significant way to detect the plagiarism which is proven by T-Test.

Choosing Mutual Fund based on Purpose, Risk Level and Inventor's Profile using Neutral Network

by Bambang H. Iswanto, Suharjito MT, and Suharjito Suharjito

Investment in financial asset these days are easily to be done. The principle in investment is hi... more Investment in financial asset these days are easily to be done. The principle in investment is higher return, higher risk. An investment with the very high return, contain very high level of risk. Otherwise if you invest in the financial asset that contain low risk level, then the expected return will be low. Therefore decision of choosing the right investment instrument is very important because it is related to risk, individual readiness in its implementation and suitability of investor profile itself. This thesis propose a machine learning method to develop a decission support system for investment manager in determining the suitable investment instrument for the individual client based on financial objective , risk level and investment period. Machine learning was choosen because it was known of its ability in recognizing the complex pattern based on learning process using the given data set. Artificial neural network multilayer perceptron (MLP) with two layers could be used as a classificator model to predict the mutual fund investment type that was suitable for investor based on investment purpose , risk level and investor profile. Experiment using 2-layers artificial neural network, 9 unit input, 16 hidden units and 4 output units that was trained with 50 data points consists of 9 vestor input component and 4 vector target components. The architecture and the data in the network could done the classification with accuration 93.33 percents.

PREDICTION OF RESEARCH TOPICS USING COMBINATION OF MACHINE LEARNING AND LOGISTIC CURVE

Prediction of future research topics is usually undertaken, especially by policy makers in resear... more Prediction of future research topics is usually undertaken, especially by policy makers in research environment, to determine the focus of future studies. Analysis of future research topics by using a numerical approach based on scientific publications or patents have been previously done by several researchers mostly by classifying the topics based on the similarity of the documents. Meanwhile, time series analysis using statistical techniques, machine learning and logistic curves has also been conducted but for areas other than research topics. In this study, we will perform time series analysis to predict the topic of research by employing combination of machine learning approaches such as Neural Network, Extreme Learning Machine and Support Vector Machine. The prediction result is then finally refined by logistic curve. The dataset used in this study is a research report on Bioinformatics from Microsoft Research and NCBI (National Center for Biotechnology Information), over the past 30 years. Experimental result indicates that the combination of machine learning approaches and logistic-curve may improve the prediction accuracy. In addition, the emerging topic of the same dataset can be predicted with a relatively high average precision. Based on the dataset used, some emerging topics predicted for the next three years are 'Sequence Alignment', 'Index Terms', 'Secondary Structure', 'Nucleic Acid', and 'Saccharomyces cerevisiae'

Implementation of Classification Technique in Web Usage Mining of Banking Company

—The goals of this paper is to implement classification technique in web usage mining to a bank c... more —The goals of this paper is to implement classification technique in web usage mining to a bank company that can help the company to identify web performance issue. Web usage mining consists of three phases: data preprocessing, pattern discovery, and pattern analysis. In pattern discovery phase, we propose to use classification technique with k-nearest neighbor algorithm implemented with standardized Euclidean distance to classifying frequent access pattern. The result shows that the k-nearest neighbor algorithm can be implemented in web usage mining and can help company to find interesting knowledge in web server log.

Modeling Software Effort Estimation Using Hybrid PSO-ANFIS

— Accurate estimating software development effort is essential in effective project management pr... more — Accurate estimating software development effort is essential in effective project management processes such as budgeting, project planning and control. To achieve an accurate estimate some algorithmic estimation techniques proposed to eliminate or reduce inaccuracies estimation. COCOMO is a parametric model used to estimate software effort. However, so far no model has proven successful to effectively and consistently predict software effort. Parametric models are considered vulnerable when faced with the problem of non-linearity of the complex in the parameters. In recent years, some estimation technique appears using intelligent systems to predict software effort. This study uses a model Neuro-fuzzy optimized with PSO to get the right model to improve the estimation effort at NASA dataset software project. Parameter cost driver, consisting of 17 feature COCOMO will then be optimized using PSO techniques to get a better prediction accuracy. Furthermore, the results of the optimization will be trained in using the algorithm to get a prediction Neuro-fuzzy effort. The performance of the proposed estimation model will be evaluated with some other intelligent system model parameters to evaluate several criteria such as Mean Standard Error (MSE), Mean Magnitude of Relative Error (MMER), and Level Prediction (Pred). The model that best shows the error rate MSE and MMER lowest to highest Pred.

Mobile Expert System Using Fuzzy Tsukamoto for Diagnosing Cattle Disease

Handling reproductive disorders in the cattle farming businesses is still ineffective due to a la... more Handling reproductive disorders in the cattle farming businesses is still ineffective due to a lack of information about the treatment of cattle, which has led to a significant decline in meat production. To reduce the impact of reproductive disease, it is necessary to perform detection and early treatment to prevent significant losses and a wider spread of the disease. This article explains the application of an expert system that provides a means of consultation imitating the reasoning process of an expert in solving complex problems concerning the health of cows reproduction. We will apply Fuzzy Tsukamoto method to help diagnose the level of risk of disease in cattle based on six clinical symptoms. The result of this research is a mobile expert system that will conclude the level of risk of endometritis disease in cattle. This mobile application is developed based on android, for ease of use, and can be used by farmers in making the diagnosis by themselves. The validation results of this expert system show that the system is able to determine the level of risk of endometritis for cow's reproduction disease. Abstract Handling reproductive disorders in the cattle farming businesses is still ineffective due to a lack of information about the treatment of cattle, which has led to a significant decline in meat production. To reduce the impact of reproductive disease, it is necessary to perform detection and early treatment to prevent significant losses and a wider spread of the disease. This article explains the application of an expert system that provides a means of consultation imitating the reasoning process of an expert in solving complex problems concerning the health of cows reproduction. We will apply Fuzzy Tsukamoto method to help diagnose the level of risk of disease in cattle based on six clinical symptoms. The result of this research is a mobile expert system that will conclude the level of risk of endometritis disease in cattle. This mobile application is developed based on android, for ease of use, and can be used by farmers in making the diagnosis by themselves. The validation results of this expert system show that the system is able to determine the level of risk of endometritis for cow's reproduction disease.

Sign Language Recognition Application Systems for Deaf-Mute People: A Review Based on Input-Process-Output

Sign Language Recognition is a breakthrough for helping deaf-mute people and has been researched ... more Sign Language Recognition is a breakthrough for helping deaf-mute people and has been researched for many years. Unfortunately, every research has its own limitations and are still unable to be used commercially. Some of the researches have known to be successful for recognizing sign language, but require an expensive cost to be commercialized. Nowadays, researchers have gotten more attention for developing Sign Language Recognition that can be used commercially. Researchers do their researches in various ways. It starts from the data acquisition methods. The data acquisition method varies because of the cost needed for a good device, but cheap method is needed for the Sign Language Recognition System to be commercialized. The methods used in developing Sign Language Recognition are also varied between researchers. Each method has its own strength compare to other methods and researchers are still using different methods in developing their own Sign Language Recognition. Each method also has its own limitations compared to other methods. The aim of this paper is to review the sign language recognition approaches and find the best method that has been used by researchers. Hence other researchers can get more information about the methods used and could develop better Sign Language Application Systems in the future. Abstract Sign Language Recognition is a breakthrough for helping deaf-mute people and has been researched for many years. Unfortunately, every research has its own limitations and are still unable to be used commercially. Some of the researches have known to be successful for recognizing sign language, but require an expensive cost to be commercialized. Nowadays, researchers have gotten more attention for developing Sign Language Recognition that can be used commercially. Researchers do their researches in various ways. It starts from the data acquisition methods. The data acquisition method varies because of the cost needed for a good device, but cheap method is needed for the Sign Language Recognition System to be commercialized. The methods used in developing Sign Language Recognition are also varied between researchers. Each method has its own strength compare to other methods and researchers are still using different methods in developing their own Sign Language Recognition. Each method also has its own limitations compared to other methods. The aim of this paper is to review the sign language recognition approaches and find the best method that has been used by researchers. Hence other researchers can get more information about the methods used and could develop better Sign Language Application Systems in the future.

Evaluation of Wavelet Fusion Method On Land Cover Classification In Bodetabek Area, Indonesia

— This paper aims to present an evaluation of wavelet fusion method on land cover classification ... more — This paper aims to present an evaluation of wavelet fusion method on land cover classification task. Wavelet fusion is one of the pan-sharpening methods which combines the higher spatial resolution panchromatic image with the lower resolution multispectral image to create high resolution fused image. Data fusion using multispectral and high spatial resolution panchromatic images are useful for improving classification accuracy. The study area of our research is Bodetabek (Bogor, Depok, Tangerang, and Bekasi) area, Indonesia. Different wavelet bases (Haar, Db2 to Db6, Coif1 to Coif5, and Sym1 to Sym5) were examined to determine the best basis for the data fusion process. This study also examined the effect of wavelet decomposition level to the spatial and spectral quality of the fused image. The experimental results on LANDSAT data show that the best basis for wavelet fusion is Coif5. The classification accuracy assessment on different wavelet decomposition level fused image also demonstrates that the higher wavelet decomposition, the higher spatial quality of the fused image. Although the spectral quality was degraded as the wavelet decomposition level increased, the classification accuracy assessment results show that higher wavelet decomposition level yields better overall classification accuracy (96.28% for eight decomposition level vs 82.77% for two decomposition level).

Online Transaction Processing (OLTP) Performance Improvement Using File-Systems Layer Transparent Compression

In this research, we use three Swingbench OLTP benchmark scenarios to examine that three compress... more In this research, we use three Swingbench OLTP benchmark scenarios to examine that three compression algorithm in ZFS file-systems namely LZ4, LZJB and ZLE can improve OLTP database performance. Beside the database performance, we also compare how much storage can be saved, impact to maximum response time, and the increase of CPU utilization from the three compression algorithms. The acquired data were then be analyzed using Analytic Hierarchy Process to find out the highest ranking compression in terms of benefits and benefits to cost ratio. The result indicates that LZJB achieved the highest performance improvement, LZ4 achieved the highest storage saving and ZLE achieved the smallest CPU utilization overhead. The safest algorithm that did not experience any reduction in database performance or increase of maximum response time in this research is LZJB.