http://dlib.iit.ac.lk/xmlui/handle/123456789/509 2026-04-08T13:04:07Z http://dlib.iit.ac.lk/xmlui/handle/123456789/1611 Suggesting waypoints to autonomous surface vessels (ASV) for effective emergency response using machine learning K.L.U.D, Lekamge, With the advancements in the maritime industry, which delivers almost 90% of the world trade, the frequency of maritime activities has drastically increased resulting a major concern in maritime safety. According to the latest accident investigation publication by European Maritime Safety Agency (2018), there had been a total of 10,384 maritime accidents reported causing 297 deaths, 273 very serious casualties and 127 ships sunk during the period from 2015 Q1 to 2018 Q2 within the European Union only. A recent research by Zhang and Li (2017) has discovered that a significant 30% of maritime accidents are caused due to bad weather conditions like sea storms which was further researched by Goerlandt et al. (2016) and Bitner-Gregerse et al. (2016) and strong winds caused due to high turbulence and high waves as discussed by Maritime Injury Center (2019). These deaths and casualties would have been minimized if there was a mechanism for efficient emergency response as discussed by Wróbel et al. (2017). ASVs have been used for several disaster mitigation and recovery operations in hurricanes such as Wilma (2005), Ike (2008) and the Tōhoku earthquake and tsunami (2011) according to Xiao et al. (2017). Therefore, ASVs could be used for emergency response since they are comparatively cheap and safe to be deployed on to hazardous zones in the deep sea since they have long term marine presence because they are mostly powered by wave energy, solar energy and wind energy as discussed by Meinig et al. (2015) and Zhou et al. (2015). A more efficient way for emergency response by the ASV would be, the ability to predict a location where there is a possibility for an accident to take place and position itself such that it could effectively respond to the emergency. Hence the author is proposing an optimal solution using machine learning techniques to suggest the waypoints to ASVs for effective emergency response on human operated surface vessels 2020-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/1610 Deep learning for ECG classification B.H.N, Sudila, The global number one deadliest cause is Cardiovascular diseases (CVDs) and it is responsible for 17.9 million global deaths annually. CVDs are a group of disorders of the heart and blood vessels and include coronary heart disease, cerebrovascular disease, rheumatic heart disease, and other conditions(Cardiovascular diseases (CVDs), 2017). Most of the CVDs are related to heartbeats and an arrhythmia is a cause with the pattern of the rhythm of the heartbeat. It means that the electrical signals responsible for heartbeats are not having smooth functionality. Irregular heartbeat signals may cause racing or fluttering heart(Different heart diseases, 2018). A cardiologist has to spend a considerable amount of time to examine ECG signals specially when the signals are beyond 30 minutes or more. ECGs are sometimes taken over a span of 48 hours. In these particular scenarios, it’s a time-consuming job and this worsens with the increasing number of patients. In this research, the main focus is to develop an open source decision support framework to classify the arrhythmias of lengthy ECG signals using a Convolutional based Deep Neural Network along with the highly regarded MIT-BIH arrhythmia dataset to train and validate the deep learning approach. The automated system considers 13 arrhythmias in classification and all of it is made generic and not-patient specific. The system is able to classify 30 minute long ECGs well within 1 minute. Extensive research and visualization methods were adopted to analyze and understand the dataset. This led to the exact numbers of data points available for each arrhythmia. The results clearly indicated that the dataset is highly unbalanced. Therefore, as an approach to solve the issue of unbalanced dataset, GANs was used in generating datasets for the arrhythmia classes with low data to increase its sensitivity and specificity values. In addition, three other high-level architectures including LSTM, DNN and CNN were tested for the same dataset before concluding on Convolutional based Deep Neural Network and their results are mentioned 2020-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/1609 Know your drug (KYD) drug supply chain management system for counterfeit drug using a hybrid block chain system N.L.S, Sithara, Counterfeit drugs are the one of most serious problems in pharmaceutical industry nowadays. According to health reports of world health organization and the health research funding organizations, nearly 10-30% of the drugs are fake in developing countries. This problem is most prevalent in such areas like drug surveillance and regulation needs to be improved and where drugs are in high demand, but mostly they remain unaffordable. When comparing with traditional drugs theses counterfeit drugs can produce various side effects to human health. Developing country like Sri Lanka also doesn’t have a proper way to track and trace drugs or to protect from this threat using any other available methods. Because counterfeiters can inject their counterfeit drugs in to drug supply chain wherever it’s vulnerable within the chain. So, identifying this problem as a timely need author has chosen to research and develop a proper system for counterfeit drugs problem by overcoming drawbacks of existing systems. This project, Know Your Drug system is designed and implemented as more transparent and trustful way to track and trace the drugs and secure the drug supply chain from counterfeiters. Implemented supply chain will overcome the trust barriers in the existing drug supply chains where an administrator or any kind of third-party intruder cannot exploit the system. Author has used a hybrid-blockchain system to approach and give solution for these conditions. Furthermore, it has given a solution for data tampering issues facing when using the private blockchains. Though, Know Your Drug system has developed more towards general usage where manufactures, distributors, wholesalers and retailers come in to play, this research paper will discuss how it is used to track and trace counterfeit drugs in Sri Lanka according to Sri Lankan regulations and systems for the drug supply chain. 2020-01-01T00:00:00Z http://dlib.iit.ac.lk/xmlui/handle/123456789/1608 FOODY - Food recipe recommendation system using sentiment analysis based neural network W.H.R.M, Wickramanayaka, Present world has seen rapid growth online review sites with a multitude of platforms available such as yummly.com, delish.com, foodnetwork.com, allrecipies.com and more food lovers tend to do searching recipes throughout recipe websites, proving that the reviews are significantly influential on food cooking process. However, reading, summarizing and categorizing hundreds of reviews by their subject matter is a time consuming and a tedious task considering the current day busy lifestyles. And also, it is challenging to understand the trends and summaries the general opinion of food lovers. Because, the real insights are hidden in the reviews used by the food recipe searchers, due to the unstructured nature of the texts. Foody is a recipe recommendation system which will use concepts like sentiment analysis, word embedding, web crawling and ranking to recommend the best salon for the user’s required service by analyzing reviews by customers 2020-01-01T00:00:00Z