Akram Pasha

Visvesvaraya Technological University, Compute Science and Engineering, Faculty Member

Followers

Following

Public Views

Aswani Kumar Cherukuri

VIT University

Hanne Thomas

FHNW

Paul Tobin

Dublin Institute of Technology

Mehmet Fatih Amasyali

Yildiz Technical University

Fabio Cuzzolin

Oxford Brookes University

Roshan Chitrakar

Nepal College of Information Technology

Lev Manovich

Graduate Center of the City University of New York

Rigas Kotsakis

Aristotle University of Thessaloniki

Ferhat Bozkurt

Ataturk University

Lyudmila S Mihaylova

The University of Sheffield

Interests

Uploads

Papers by Akram Pasha

Insights of Mathematics for Big Data

International Journal of Engineering and Advanced Technology

Computer Science can be considered as one of the extensions made to the pure mathematical science... more Computer Science can be considered as one of the extensions made to the pure mathematical sciences that exhibit the design and development of many mathematical models to solve various engineering problems. Data storage and data processing are the two major operations that are primarily focused by any computational model while solving a problem. Mathematical modelling has been helped in producing the various computational models across several problems that are found in the field of computer science. Among many problems that are found in the area of computer science, data science and big data have recently geared up to solve many business oriented problems that are purely based on data analytics to enhance the profit by taking critical business decisions. Data Scientists and mathematicians are found to have a skeptical understanding or too little collaboration either in knowing the mathematical concepts behind big data technologies, or too little knowledge of applications of mathemat...

Comparative Study of Multiple Machine Learning Algorithms for Students’ Performance Data for Job Placement in University

International Journal of Engineering and Advanced Technology

Download

Multiple Machine Learning Classifiers for Student’s Admission to University Prediction

International Journal of Engineering and Advanced Technology

A Critical Study : Secure Gateway in Payment System for Electronic Transaction

In this paper the attempt of studying a Secure Gateway in payment system for Electronic transacti... more In this paper the attempt of studying a Secure Gateway in payment system for Electronic transaction is made. The current research and development shows that the electronic payment system for such an electronic transaction is to be secure for participants such as Payment Gateway Server, Bank Servers and Merchant Servers, on Internet. The security architecture of such systems are designed by using various Security Protocols and Techniques, not only to safeguard but eliminate the fraud that occurs in such a transaction with stolen credit card/debit card payment information and customer information. Electronic commerce involves the tremendous exchange of some form o f money for trade of goods and services over the Internet, and it is evident from the studies that the Internet is an insecure and unreliable media in many ways in such a trade. The primary goal of this paper is to review the asymmetric key crypto-system methodology that uses Security Protocol, the Secure Communication Tunne...

Download

Efficient Sensory Data Transformation: A Big Data Approach

Intelligent Systems Applications in Software Engineering

Big Data Analytics has immensely solved complex problems involving massive and complex data. Data... more Big Data Analytics has immensely solved complex problems involving massive and complex data. Data filtering plays a vital role in helping the analysis processes to perform analytics with ease and in precise form. The data variety and veracity are some of the major problems that add enough complexity to data, creating the overall hindrance to an effective big data analytics process. The proposed work targets the data staging phase in a big data classification to tackle variety and veracity problems found in massive sensory data. The study adopts a novel approach for data storage, then designs and implements a simple algorithm to perform data transformation. The concept of cloud storage-bucket is used for effective storage and transformation of sensory data. Such an analytical approach is proven to retain the capability of an effective and faster data transformation. The algorithm performs conversion of unstructured data to semi-structured data in the first stage, then converts the semi-structured data to structured data in the second stage, and finally stores the resulting structured data into virtually localized distributed storage. The outcome of this study offers faster response time and higher data purity for data transformation process of data staging phase in any big data analytics application.

Well-Calibrated Probabilistic Machine Learning Classifiers for Multivariate Healthcare Data

International Journal of Advanced Research in Computer Science

The healthcare applications frequently collect and store the patient data (mostly multivariate) t... more The healthcare applications frequently collect and store the patient data (mostly multivariate) to examine the history of the treatment and thereby enhance the effectiveness of treatment. The efficient treatment to the patient depends on the performance of the machine learning models used for analytics tasks of patient data. It is convenient to have a machine learning classification model in a healthcare application to predict the probability of an observation belonging to each possible class rather than predicting a class value directly for any disease classification problem. Such predicted probabilities are required to be calibrated to assist the overall support and confidence of any machine learning classification model used in many healthcare applications. In this paper, the predicted probabilities are studied to diagnose and i mprove the calibration of models used for probabilistic classification. The general performance of selected classification models on the two latest wart skin disease treatment data is also reported.

Download

Bio-inspired dimensionality reduction for Parkinson’s disease (PD) classification

Health Information Science and Systems

Given the demand for developing the efficient Machine Learning (ML) classification models for hea... more Given the demand for developing the efficient Machine Learning (ML) classification models for healthcare data, and the potentiality of Bio-Inspired Optimization (BIO) algorithms to tackle the problem of high dimensional data, we investigate the range of ML classification models trained with the optimal subset of features of PD data set for efficient PD classification. We used two BIO algorithms, Genetic Algorithm (GA) and Binary Particle Swarm Optimization (BPSO), to determine the optimal subset of features of PD data set. The data set chosen for investigation comprises 756 observations (rows or records) taken over 755 attributes (columns or dimensions or features) from 252 PD patients. We employed MaxAbsolute feature scaling method to normalize the data and one hold cross-validation method to avoid biased results. Accordingly, the data is split in to training and testing set in the ratio of 70% and 30%. Subsequently, we employed GA and BPSO algorithms separately on 11 ML classifiers (Logistic Regression (LR), linear Support Vector Machine (lSVM), radial basis function Support Vector Machine (rSVM), Gaussian Naïve Bayes (GNB), Gaussian Process Classifier (GPC), k-Nearest Neighbor (kNN), Decision Tree (DT), Random Forest (RF), Multilayer Perceptron (MLP), Ada Boost (AB) and Quadratic Discriminant Analysis (QDA)), to determine the optimal subset of features (reduction of dimensionality) contributing to the highest classification accuracy. Among all the bio-inspired ML classifiers employed: GA-inspired MLP produced the maximum dimensionality reduction of 52.32% by selecting only 359 features and delivering 85.1% of the classification accuracy; GA-inspired AB delivered the maximum classification accuracy of 90.7% producing the dimensionality reduction of 41.43% by selecting only 441 features; And, BPSO-inspired GNB produced the maximum dimensionality reduction of 47.14% by selecting 396 features and delivering the classification accuracy of 79.3%; BPSOMLP delivered the maximum classification accuracy of 89% and produced 46.48% of the dimensionality reduction by selecting only 403 features.

Multiple Machine Learning Classifiers for Student’s Admission to University Prediction

International Journal of Engineering and Advanced Technology (IJEAT), 2019

Data is the most important asset for any organization which is further processed to produce usefu... more Data is the most important asset for any organization which is further processed to produce useful information. Machine Learning and Big Data techniques are widely used for industrial sectors to generate useful patterns helpful for earning more profits and expand businesses. From the past few years, a lot of research works have been done by using Big Data techniques on educational data for improvement in Education System. Machine Learning and Big Data can be useful for predicting the students' admission, performance of teaching, performance of a student, identifying the group of students of similar behavior. However, the manual process of record checking is time consuming, tedious, and error prone; due to the inherent volume and complexity of data. In this study, the combination of linear and non-linear machine learning algorithms; Logistic Regression, Decision Tree, k-NN, and Naïve Bayes have been chosen to perform prediction of the target class for an unseen observation by polling. As the models built in this work are predicting the likelihood of a student taking up the admission into any university based on the student data collected by any marketing agency, the combined models are collectively called as the Admission Predictor. The administrative officials of any academic institution can use this kind of an application to explore and analyze the patterns that are affecting the student admission and come up with enhanced strategies to improve admission. Such an application not only plays a vital role in administration, but also help the management in reformulating the marketing criteria for overall development of academic institution.

Download

Comparative Study of Multiple Machine Learning Algorithms for Students’ Performance Data for Job Placement in University

International Journal of Engineering and Advanced Technology (IJEAT), 2019

In the era of data evolution, many organizations have taken the lead in storing the data in huge ... more In the era of data evolution, many organizations have taken the lead in storing the data in huge data repositories. Analysis of data comes with several challenges since the time the data is captures till the insights are inferred from the data. Accentuating the accuracy of data analysis is of paramount importance as many critical decisions are totally dependent on the outcomes of the analysis. Machine learning has been found as the most effective and most preferred tool in the literature for in-memory data analytics. Universities mostly collect the statistical data related to the students that is only either used quantitatively or sparsely analyzed to gain the insights that could be useful for the authorities to enhance the percentage of placements in campus drives held through early analysis of such data accurately. The work proposed in this paper formulates the problem of predicting the likelihood of a student getting placed in a company as a binary classification problem. Then it makes an effort to train and perform the empirical study of following multiple machine learning algorithms with the placement data; Logistic Regression, Naïve Bayes, Support Vector Machine, K-Nearest Neighbor and Decision Tree. The machine learning classification models are built to predict the probabilities of a student getting placed in a company based on the student's academic scores, achievements, work experience (internship), and many other relevant features. Such an analysis helps the university authorities to dynamically create plans to enhance the unlikely students to be placed in a company participating in the campus recruitment held in the university. To improve these models and to avoid the models from overfitting to the training data, strategies like K-Fold cross-validation is applied for various values of k. The machine learning models selected are also compared for its efficiency by employing the supervised and unsupervised feature extraction techniques such as PCA and LDA. The Decision Tree model with K as 10 for cross-validation and PCA has outperformed all the other models producing the accuracy of 72.83% with satisfactory support and recall during experimentation. The application focuses on the targeted group of students, to eventually improve the probability of students getting placed during campus recruitment drives held in the university.

Download

Machine learning Classifiers for Credit Card Fraud Detection: A Brief Survey

International Journal of Computer Sciences and Engineering, 2019

Utilization of credit cards encourages individuals to buy products online via the Internet. Indiv... more Utilization of credit cards encourages individuals to buy products online via the Internet. Individuals tend to do much of purchasing online or offline by utilizing the credit card facility provided by the bankers to their customers. Credit cards have turned out to be the most prominent facility available to the people around the globe to encourage paperless trades at an enormous speed. Whenever any such trade happens in exchanges or net marketing by using a paperless framework, it is subjected under high risk of fraudulent transactions due to many pitfalls in the security system of the usage of credit cards on the networks. This paper presents a brief survey of important and basic linear and non-linear machine learning algorithms that are focused to predict the fraudulent transactions by studying the patterns present in the credit card transactional datasets. The authors provide the methodology of Random Forest (RF), Support Vector machine (SVM) and Artificial Neural Network (ANN) classifiers to accurately classify whether a unseen credit card transaction is fraudulent or not.