http://dlib.iit.ac.lk/xmlui/handle/123456789/699 2026-05-04T07:40:19Z http://dlib.iit.ac.lk/xmlui/handle/123456789/793 Sports strokes identification and classification an videos using spatio temporal feature extraction Bandara, Herath Mudiyanselage Ishara Sandun " In this research, a novel approach for stroke classification in sports using video based technologies is introduced. With the proposed approach strokes can be identified and classified in near real time from streaming videos. The approach uses a windowing approach to identify stroke play events and then uses human motion modeling and analysis (HMMA) techniques to extract spatio-temporal features from videos. The spatio-temporal time series datasets are used with deep neural networks (LSTM variants) for classification of strokes. Proposed approach can be used with multiple sports like cricket, tennis, badminton, table tennis, baseball etc. In this thesis, implementation of the proposed approach with two different sports (cricket and tennis) is discussed. Three LSTM variants (LSTM network, Bi-directional LSTM network, CNN- LSTM network) have been tested in order to find the most suitable neural network. All trained classifiers in both case studies achieved over 95% average accuracies with the proposed approach. Bi-directional LSTM network has achieved the best accuracy for the classification task in both case studies. The proposed approach has the potential to be useful in areas like augmented coaching, television broadcasting, sports analytics etc. Future enhancements to this proposed approach will include use of the proposed approach for augmented coaching with mobile technologies." 2021-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/792 "Predicting the impact of policy decisions on the reduction of COVID-19 cases detected using historical data " Amjadh, Mohamed Ifthikar Mohamed Abstract Since the beginning of the year 2020, the world has been taken over by the claws of the Covid-19 pandemic. Since then the lives of the people living on the earth have not been the same. Large numbers of people have been dieting on a daily basis and economies around the world have been collapsing. In this crisis, the best thing people can do to safeguard themselves is to get the vaccination and ensure proper health guidelines are followed on a day to day basis. Based on the nature of the virus, a person can be easily recovered given that proper medical care is provided. Around the world it is evident that the amount of deaths are skyrocketing whenever the health services of a country are overloaded. Governments therefore take a lot of measures to ensure that the rate of infections does not increase in a drastic manner. This is also famously known as flattening the curve. In order to flatten the curve the governments impose and enforce different health measures from curfews to mandating wearing masks. When imposing these restrictions, it is imperative to ensure that unnecessary damage is not done to the economy. This means that the lockdowns need to be done in a scientific manner. The goal of this research is to create a tool that can be used to determine the effectiveness of a lockdown beforehand so that only the necessary restrictions can be imposed in order to reach the target virus control. By the conclusion of the research we were able to come up with a solution using a multivariate time series forecasting model based on a bi-directional LSTM. The created solution was 91% accurate in forecasting the future of a virus spread based on lockdown details being imposed. 2021-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/791 Premium prediction using risk assessment to generate smart contracts for the health insurance sector Wijendra, Kanishka Lahiru " Insurance is a mode of transferring the risk and it comprises of several products. Health insurance sector being the foundation of this study, has various activities that are integrated in its process. Many of those being manual activities with lengthy processes, often leads to inefficiencies. Through this study, a basic solution will be developed to overcome the inefficiencies in the health insurance process. An automation of this process will be developed as a solution. Smart contract technology of the blockchain paradigm will be used in this automation due to its characteristic of immutability of transactions. The decentralized framework of blockchain networks will allow the solution to be secured. The smart contracts will be generated to replace the traditional health insurance policies with this application. Replacement of traditional health insurance policies with an automated approach will help in enabling faster insurance claim processes and avoid inconsistencies. Process automation will require the insurance premium to be predicted prior to generating a smart contract. Several predictive algorithms will be evaluated to arrive at best fit model. Regression approaches such as multiple linear regression, lasso regression, ridge regression, regression tree method and gradient boosting will be evaluated through this study. Fitted models will be evaluated using evaluation criteria such as AIC, RMSE, and adjusted R squared. Further the best fit model will be validated using K-fold cross validation approach. This study considers gradient boosting as the best fit model for predicting health insurance premium with an accuracy of 83.7% obtained through K fold cross validation. Being a study that is focused in helping to overcome existing issues in the domain of health insurance, this will be a value addition for the insurance industry." 2021-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/790 District level forecasting of OPD patient visits in critical districts in Sri Lanka - Case Study Fernando, Wannaku Waththa Mitiwaduge Dewni Yasara " Insurance is a mode of transferring the risk and it comprises of several products. Health insurance sector being the foundation of this study, has various activities that are integrated in its process. Many of those being manual activities with lengthy processes, often leads to inefficiencies. Through this study, a basic solution will be developed to overcome the inefficiencies in the health insurance process. An automation of this process will be developed as a solution. Smart contract technology of the blockchain paradigm will be used in this automation due to its characteristic of immutability of transactions. The decentralized framework of blockchain networks will allow the solution to be secured. The smart contracts will be generated to replace the traditional health insurance policies with this application. Replacement of traditional health insurance policies with an automated approach will help in enabling faster insurance claim processes and avoid inconsistencies. Process automation will require the insurance premium to be predicted prior to generating a smart contract. Several predictive algorithms will be evaluated to arrive at best fit model. Regression approaches such as multiple linear regression, lasso regression, ridge regression, regression tree method and gradient boosting will be evaluated through this study. Fitted models will be evaluated using evaluation criteria such as AIC, RMSE, and adjusted R squared. Further the best fit model will be validated using K-fold cross validation approach. This study considers gradient boosting as the best fit model for predicting health insurance premium with an accuracy of 83.7% obtained through K fold cross validation. Being a study that is focused in helping to overcome existing issues in the domain of health insurance, this will be a value addition for the insurance industry." 2021-01-01T00:00:00Z