Papers by Sameerchand Pudaruth
Enhancing Stock Price Forecasting with Generative Adversarial Networks and Conformal Prediction: A Novel Approach for Quantifying Uncertainty
A Real-Time Traffic Jam Detection and Notification System Using Deep Learning Convolutional Networks
International Journal of Advanced Computer Science and Applications
Flowers are admired and used by people all around the world for their fragrance, religious signif... more Flowers are admired and used by people all around the world for their fragrance, religious significance, and medicinal capabilities. The accurate taxonomy of these flower species is critical for biodiversity conservation and research. Non-experts typically need to spend a lot of time examining botanical guides in order to accurately identify a flower, which can be challenging and time-consuming. In this study, an innovative mobile application named FloralCam has been developed for the identification of flower species that are commonly found in Mauritius. Our dataset, named FlowerNet, was collected using a smartphone in a natural environment setting and consists of 11660 images, with 110 images for each of the 106 flower species. Seventy percent of the data was used for training, twenty percent for validation and the remaining ten percent for testing. Using the approach of transfer learning, pre-trained convolutional neural networks (CNNs) such as the InceptionV3, MobileNetV2 and ResNet50V2 were fine tuned on the custom dataset created. The best performance was achieved with the fine tuned MobileNetV2 model with accuracy 99.74% and prediction time 0.09 seconds. The best model was then converted to TensorFlow Lite format and integrated in a mobile application which was built using Flutter. Furthermore, the models were also tested on the benchmark Oxford 102 dataset and MobileNetV2 obtained the highest classification accuracy of 95.90%. The mobile application, the dataset and the deep learning models developed can be used to support future research in the field of flower recognition.
Kathmandu School of Law Review, 2017
In a contextualized approach, the authors have revisited the DTAA 1 between Mauritius and India t... more In a contextualized approach, the authors have revisited the DTAA 1 between Mauritius and India to reflect to what extent RoundTripping and Treaty Shopping have an impact in the bilateral agreement between India and Mauritius. The DTAA between India and Mauritius2was signed in August 1982, and the spirit of the bilateral agreement and the negotiations, which were carried out afterwards successively, were to provide exemptions from shareholders as who have already been taxed in Mauritius should not be taxed further. However, exemptions from capital gains tax in Mauritius would also mean that tax evasion soon becomes the center of recent negotiations between the two countries with serious concerns over tax abuses, round tripping and treaty shopping. Nevertheless, although Mauritius is considered a tax haven, there are still very strong ties betweenthe two countries both historically and financially with mutual economic and financial support in a win-win situation. Indeed, Mauritius co...
The Bluetooth protocol can be used for intervehicle communication equipped with Bluetooth devices... more The Bluetooth protocol can be used for intervehicle communication equipped with Bluetooth devices. This work investigates the challenges and feasibility of developing intelligent driving system providing timesensitive information about traffic conditions and roadside facilities. The architecture for collaborative vehicle communication system is presented using the concepts of wireless networks and Bluetooth protocol. We discuss how vehicles can form mobile ad-hoc networks and exchange data by the on-board Bluetooth sensors. The key design concepts of the intelligent driving service infrastructure are analyzed showing collaborative fusion of multiple positional data could give a better understanding of the surrounding traffic conditions for collaborative driving. The technical feasibility of using Bluetooth for data exchange among moving vehicles is evaluated.
International Journal of Advanced Computer Science and Applications, 2018
Lecture materials cover a broad variety of documents ranging from e-books, lecture notes, handout... more Lecture materials cover a broad variety of documents ranging from e-books, lecture notes, handouts, research papers and lab reports amongst others. Downloaded from the Internet, these documents generally go in the Downloads folder or other folders specified by the students. Over a certain period of time, the folders become so messy that it becomes quite difficult to find our way through them. Sometimes files downloaded from the Internet are saved without the certainty that they will be used or revert to in the future. Documents are scattered all over the computer system, making it very troublesome and time consuming for the user to search for a particular file. Another issue that adds up to the difficulty is the improper naming conventions. Certain files bear names that are totally irrelevant to their contents. Therefore, the user has to open these documents one by one and go through them to know what the files are about. One solution to this problem is a file classifier. In this paper, a file classifier will be used to organise the lecture materials into eight different categories, thus easing the tasks of the students and helping them to organise the files and folders on their workstations. Modules each containing about 25 files were used in this study. Two machine learning techniques were used, namely, decision trees and support vector machines. For most categories, it was found that decision trees outperformed SVM.
Theory, Methodology, Practice, 2017
This research is mainly focussed on analysing a Performance Management System (PMS) and its impac... more This research is mainly focussed on analysing a Performance Management System (PMS) and its impact on the motivational level of employees at the Ministry of Finance and Economic Development (MOFED) of the Republic of Mauritius. A survey was carried out using the convenience sampling technique to grasp the views of the employees. 200 questionnaires were distributed and the response rate was 55%. The results indicate that the main elements of a PMS process which are motivating elements for employees were missing in the organisation: regular feedback, communication and training. Thus, recommendations were made to the top management to market the concept of PMS, improve the communication strategy and feedback strategy, encourage employee participation and recognition, link PMS to pay and training and provide a supporting management who would promote organisational learning.
World Academy of Science, Engineering and Technology, International Journal of Social, Behavioral, Educational, Economic, Business and Industrial Engineering, Oct 29, 2010
In this paper, we give an overview of an online elearning tool which has been developed for kids ... more In this paper, we give an overview of an online elearning tool which has been developed for kids aged from nine to eleven years old in Mauritius for the self-study of Mathematics in order to prepare them for the CPE examination. The software does not intend to render obsolete the existing pedagogical approaches. Nowadays, the teaching-learning process is mainly focused towards the classroom model. Moreover, most of the e-learning platforms that exist are simply static ways of delivering resources using the internet. There is nearly no interaction between the learner and the tool. Our application will enable students to practice exercises online and also work out sample examination papers. Another interesting feature is that the kid will not have to wait for someone to correct the work as the correction will be done online and on the spot. Additional feedback is also provided for some exercises.
World Academy of Science, Engineering and Technology, International Journal of Social, Behavioral, Educational, Economic, Business and Industrial Engineering, Oct 29, 2010
International journal of computer applications, May 17, 2013
In this paper, a novel system is presented for the allocation of final year projects for the Comp... more In this paper, a novel system is presented for the allocation of final year projects for the Computer Science and Engineering Department at the University of Mauritius. Earlier works had concentrated only on the allocation of projects to students. The system not only performs project allocation but it also allows academics to rate projects, examiners to bid for projects they wish to examine, students to propose their own projects, students to submit project deliverables, supervisors to follow projects more closely and allows projects coordinators to have a heuristic view of the whole system. The system captures the preferences of examiners as well as students and allocates projects to them in order to maximise the number of students who gets their first choice in their preference list and to keep the load of supervisors and examiners within a reasonable range. The percentage of students who obtained their first choice is 82% on 30 projects proposed by 15 supervisors for 11 teams. The simulation results demonstrate that this new system will allow deadlines for all the different project phases to be met.
Procedia Computer Science, 2015
Name-Centric Gender Inference Using Data Analytics
In this era of globalisation and technology, determining the gender of a person from forenames ha... more In this era of globalisation and technology, determining the gender of a person from forenames has numerous applications especially in the machine translation and natural language processing fields. In this paper, we used a supervised machine learning approach to classify 10000 first names into either a male or female name. The names were manually extracted from an online telephone directory and then manually classified into an appropriate category. We obtained the highest accuracy of 88.0% when using support vector machines while the Naïve Bayes produced the lowest accuracy of 84.7%. A total of 15 features were used in this study. Traditionally, such systems have relied on a name dictionary to output the gender of forenames. However, our proposed system can predict the gender of unseen or unknown names. Furthermore, our dataset consists of names from different origins such as European, African, Arabic, Indian and Chinese, unlike previous studies which use names from one origin only.
Categorisation of Supreme Court Cases Using Multiple Horizontal Thesauri
Advances in intelligent systems and computing, Aug 22, 2015
Text classification is a branch of Artificial Intelligence which deals with the assignment of tex... more Text classification is a branch of Artificial Intelligence which deals with the assignment of textual documents to a controlled group of classes. The aim of this paper is to assess the use of a controlled vocabulary in the categorisation of legal texts. Controlled vocabularies such as the Medical Subject Headings, Compendex and AGROVOC have been proved to be very useful in the fields of biomedical research, engineering and agriculture, respectively. In this work, a number of lexicons are created for some pre-defined areas of law through an automated approach. The lexicons are then used to categorise cases from the Supreme Court into eight distinct areas of law. We then compared the performance of these lexicons with each other. We found that lexicons which have a mixture of single words and short phrases performs slightly better than those consisting simply of single words. Weights were also assigned to the terms and this had a significant positive impact on the classification accuracy. The number of words in each thesaurus was kept constant. A hierarchical classification was also attempted whereby cases were first classified into either a civil case or a criminal case. Civil cases were then further classified into company, labour, contract and land cases while criminal cases were classified into drugs, homicide, road traffic offences and other criminal offences. Our best model achieves a global accuracy of 78.9 %. Thus, we have demonstrated that it is possible to get good classification accuracies with legal cases through the use of automatically generated thesauri. This outcome of this research can become an integral part of the eJudiciary project that has already been initiated by the government. In line with the vision of the Judiciary, we are hereby in the process of creating an intelligent legal information system which will benefit all legal actors and will have a definite positive impact on the legal landscape of the Republic of Mauritius. Lawyers, attorneys and their assistants would spend less time on legal research and hence they would have more time to prepare their arguments for their case. We are optimistic in believing that this will make the whole business of providing justice more effective and more efficient through the reduction in postponement of cases and a reduction in the average disposal time of cases.
A Statistical and Machine Learning Approach for Summarising Computer Science Research Papers
International Journal of Computing and Digital Systems, Apr 16, 2023
International Journal of Computing and Digital Systems, Jan 31, 2023
Nowadays, many people are unaware of the benefits of fruits and vegetables which has resulted in ... more Nowadays, many people are unaware of the benefits of fruits and vegetables which has resulted in their reduced consumption. This has inevitably led to a rise in diseases such as obesity, high blood pressure and heart diseases. To this end, we have developed FruVegy which is an android app which can automatically identify fruits and vegetables and then display its nutritional values. The app can identify forty different fruits/vegetables. The app is specially targeting school students who will find it easy and fun to use and this, we believe, will increase their interest in the consumption of fruits and vegetables. Furthermore, the names of the fruits and vegetables are also available in French and in Mauritian Creole. Our dataset consists of 1600 images from 40 different fruits and vegetables. There was an equal number of images for each fruit/vegetable. To our knowledge, this is the largest dataset that currently exists in literature. Features such as shape, colour and texture were extracted from each image. Different machine learning classifiers were tested but random forest with 100 trees produced the best result with an accuracy of 90.6%. However, with TensorFlow, an average accuracy of 98.1% was obtained under different scenarios. In the future, we intend we increase our dataset and the number of features in order to achieve an even higher accuracy.
Phyllanthus phillyreifolius
Elsevier eBooks, 2020
Markov Chain Monte Carlo Methods and Evolutionary Algorithms for Automatic Feature Selection from Legal Documents
Advances in intelligent systems and computing, Oct 21, 2017
In this paper, we present three different approaches for feature selection, starting from a naive... more In this paper, we present three different approaches for feature selection, starting from a naive Markov Chain Monte Carlo random walk algorithm to more refined methods like simulated annealing and genetic algorithms. It is typical for textual data to have thousands of dimensions in their feature space which makes feature selection a crucial phase before the final classification. Classification of legal documents into eight categories was performed via a simple document similarity measure based on term frequency and the nearest neighbour concept. With an average success rate of 76.4%, the random walk algorithm not only performed better than the simulated annealing and genetic algorithms but also matched the accuracy of support vector machines. Although these methods have commonly been used for selecting appropriate features in other fields, their use in text categorisation have not been satisfactorily investigated. And, to our knowledge, this is the first work which investigates their use in the legal domain. This generic text classification framework can further be enhanced by using an active learning methodology for the selection of training samples rather than following a passive learning approach.
Development of an Incident Prioritization Model Using Fuzzy Logic to Improve Quality and Productivity in IT Support Services
Advances in intelligent systems and computing, 2019
Managing a high volume of incidents is a very complicated task for companies which provide suppor... more Managing a high volume of incidents is a very complicated task for companies which provide support services. The application support analysts as well as managers must effectively assess the level of importance of incidents during the day to better prioritize each of them. As this process is very complex and time consuming, a lot of efforts are spent in the incident prioritization activity, which is manually carried out by the first level support team and by the analysts at the start of their shift and during the workday by going through each of the incidents and determining the order on which they need to be worked on. Bad incident prioritization leads to a decrease in quality of service as analysts fail to manage customers’ expectations and this impacts productivity. To reduce this problem, a system which allows prioritization of incidents was proposed. To implement the solution, the range of factors which contributes to determine the priority of an incident was identified and a survey was conducted in multiple companies involved in ITSM to determine the importance of each of these factors. A fuzzy logic approach was formulated to determine the final priority of an incident. The results show a 19% increase in productivity and a 9% increase in quality of service.
Extraction of Roads from Remotely Sensed Images using a Multi-Angled Template Matching Technique
The extraction of road networks from satellite or aerial images has profound applications in the ... more The extraction of road networks from satellite or aerial images has profound applications in the fields of urban planning, setting up of transportation networks, disaster management, cartography and in Geographical Information Systems. In this paper, we have developed a multi-shaped and multi-angled template matching algorithm in order to extract the road network from medium and high-resolution satellite images. We used a quadruple orthogonal line filter to extract lines from four different directions. Small isolated points and edges are removed using appropriately designed clearing filters. Gaps in the road network are bridged using our edge linking algorithm, which is based primarily on the spectral property of the original image pixels. The four images are cleared again using directional clearing filters to remove broken edges that cannot be linked to the road network. Finally, the output from these four separate images are fused into a single image in order to get the final output image which represents the road network. The results obtained demonstrate the practicability of our proposed method in rural and semi-urban regions.
Using a thesaurus-based approach for the categorisation of web sites
With the increasing number of Mauritian-owned websites on the internet, the need for classificati... more With the increasing number of Mauritian-owned websites on the internet, the need for classification is becoming highly important. Our objective in this research is to classify a list of websites into seven broad categories namely education, entertainment, government, health, tourism, sports and shopping. The homepage of three hundred and nineteen websites have been used in this study. We have exploited the rich source of information (features) contained in the homepage like the meta tags, title tag, heading tags, hyperlinks, the content of the website and the domain name of the website. These information were then used to classify the websites into their most appropriate category. Several parameters like the weight applied to each feature and the keywords used to classify the websites were tuned to yield better results. The experimental evaluation revealed that the method implemented provides very high accuracy. In particularly, we obtained an accuracy of about 95% which is higher than all existing approaches considered so far in the research literature.
Uploads
Papers by Sameerchand Pudaruth