Research Publications
Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4194
This main community comprises five sub-communities, each representing the academic contribution made by SLIIT-affiliated personnel.
Browse
9 results
Filters
Advanced Search
Filter by
Settings
Search Results
Publication Open Access QPred: A Lightweight Deep Learning-Based Web Pipeline for Accessible and Scalable Streamflow Forecasting(Tech Science Press, 2026) Makumbura, R.K; Wijesundara, H; Sajindra, H; Rathnayake, U; Kumar, V; Duraibabu, D; Sen, SAccurate streamflow prediction is essential for flood warning, reservoir operation, irrigation scheduling, hydropower planning, and sustainable water management, yet remains challenging due to the complexity of hydrological processes. Although data-driven models often outperform conventional physics-based hydrological modelling approaches, their real-world deployment is limited by cost, infrastructure demands, and the interdisciplinary expertise required. To bridge this gap, this study developed QPred, a regional, lightweight, cost-effective, web-delivered application for daily streamflow forecasting. The study executed an end-to-end workflow, from field data acquisition to accessible web-based deployment for on-demand forecasting. High-resolution rainfall data were recorded with tipping-bucket gauges and loggers, while river water depth in the Aglar and Paligaad watersheds was converted to discharge using site-specific rating curves, resulting in a daily dataset of precipitation, river water level and discharge. Four DL architectures were trained, including vanilla Long Short-Term Memory (LSTM), stacked LSTM, bidirectional LSTM, and Gated Recurrent Unit (GRU), and evaluated using Nash-Sutcliffe Efficiency (NSE), Coefficient of Determination (R2), Root-Mean-Square-Error-Standard-Deviation Ratio (RSR), and Percentage Bias (PBIAS) metrics. Performance was watershed-specific, as the vanilla LSTM demonstrated the best generalisation for the Aglar watershed (R2 = 0.88, NSE = 0.82, RMSE = 0.12 during validation), while the GRU achieved the highest validation accuracy in Paligaad (R2 = 0.88, NSE = 0.88, RMSE = 0.49). All models achieved satisfactory to excellent performance during calibration (R2 > 0.91, NSE > 0.91 for both watersheds), demonstrating strong capability to capture streamflow dynamics. The highest performing models were selected and embedded into the QPred application. QPred was developed as a lightweight web pipeline, utilising Google Colab as the primary execution environment, Flask as the backend inference framework, Google Drive for artefact storage, and Ngrok for secure HTTPS tunnelling. A user-friendly front end utilises range sliders (bounded by observed minima and maxima) to gather inputs and provides discharge data along with metadata, thereby enhancing transparency. This work demonstrates that accurate, context-aware deep learning models can be delivered through low-cost, web-based platforms, providing a reproducible and scalable pipeline for hydrological applications in other watersheds and for practitioners. CopyrightPublication Embargo Comparison of ARIMA and LSTM in Forecasting the Retail Prices of Vegetables in Colombo, Sri Lanka(IEEE, 2022-12-09) Fonseka, D.D; Karunasena, AIdentification of vegetable price trends is important to make better decisions in the production and market. Due to several factors, including seasonality, perishability, an imbalanced supply-demand market, customer choice, and the availability of raw materials, vegetable prices fluctuate quickly and are highly unstable. In this study price prediction was concluded using two models ARIMA and LSTM with retail price data for Cabbage, Carrot, and Green beans in Colombo from 2009 to 2018. According to the decision criteria of RMSE and MAPE, the LSTM model is superior to the ARIMA model in predicting the retail prices of vegetables. There were no studies have focused on predicting prices with novel technology in the Sri Lankan vegetable market. Hence the results of this study can be used to build an advanced forecasting model by the government and decision-makers in agriculture in Sri Lanka.Publication Open Access Cryptocurrency Price Prediction: A Comparative Study using LSTM, GRU and Stacking Ensemble Algorithm for Time Series Forecasting(SLIIT, 2022-02-11) Ashikul Islam, M. DTechnology has significantly reshaped how humans interact with their tangible and intangible surroundings. Cryptocurrency is considered to be one of the most recent technological inventions which revolutionized how we perceive currencies and their functionality. It has become popular because of its safety, security and anonymity. However, volatility remains one of the major issues with cryptocurrencies to this day. Therefore, the primary aim of this paper is to develop LSTM (Long ShortTerm Memory), GRU (Gated Recurrent Units) and a Stacking Ensemble Learning algorithm that efficiently predicts the price of a cryptocurrency for a given period of time. The predictions are then observed and analysed to determine the comparative performance of the said algorithms.Publication Open Access Agro-Genius: Crop Prediction Using Machine Learning(https://ijisrt.com/agrogenius-crop-prediction-using-machine-learning, 2019-10) Gamage, M. P. A. W; Kasthurirathna, D; Paresith, M. M; Thayakaran, S; Suganya, S; Puvipavan, PThis paper present a way to aid farmers focusing on profitable vegetable cultivation in Sri Lanka. As agriculture creates an economic future for developing countries, the demand of modern technologies in this sector is higher. Key technologies used for this problem are Deep Learning, Machine Learning and Visualization. As the product, an android mobile application is developed. In this application the users should input their location to start the prediction process. Data preprocessing is started when the location is received to the system. The collected dataset divided into 3 parts. 80 percent for training, 10 percent for testing and 10 percent for validation. After that the model is created using LSTM RNN for vegetable prediction and ARIMA for price prediction. Finally, for given location profitable crop and predicted future price of vegetables are shown in the application. Other than the prediction, optimizing for multiple crop sowing according to the user requirements and visualizing cultivation and production data on map and graphs are also given in the application. This paper elaborates the procedure of model development, model training and model testing.Publication Embargo Comparative analysis of the application of Deep Learning techniques for Forex Rate prediction(2019 1st International Conference on Advancements in Computing (ICAC), SLIIT, 2019-12-05) Aryal, S.; Nadarajah, D.; Kasthurirathna, D.; Rupasinghe, L.; Jayawardena, C.Forecasting the financial time series is an extensive field of study. Even though the econometric models, traditional machine learning models, artificial neural networks and deep learning models have been used to predict the financial time series, deep learning models have been recently employed to do predictions of financial time series. In this paper, three different deep learning models called Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN) and Temporal Convolution Network (TCN) have been used to predict the United States Dollar (USD) to Sri Lankan Rupees (LKR) exchange rate and compared the accuracy of the models. The results indicate the superiority of CNN model over other models. We conclude that CNN based models perform best in financial time series prediction.Publication Embargo Exploiting Multivariate LSTM Models with Multistep Price Forecasting for Agricultural Produce in Sri Lankan Context(2020 2nd International Conference on Advancements in Computing (ICAC), SLIIT, 2020-12-10) Navaratnalingam, S.In Sri Lanka agricultural produces possess a large supply which involves various stakeholders and thus, fluctuation of the agricultural produce prices has a direct impact on the purchasing decisions of the consumer. So, the main purpose of this study is to address the problem faced by the consumer due to poor awareness of price fluctuation which consequently astonish the consumers and hinder them from making better purchasing decisions. The research study is being specially developed in a way to adapt the Sri Lankan agricultural consumer market that is mainly based on Pettah and Dambulla trade centers. As the study we exploited different types of LSTM model with multivariate inputs along with the different combination of multistep models. The result of the study reveals that better performance was obtained for the multivariate CNN LSTM model with encoder decoder multistep model which provided an average RMSE of 19.46 Sri Lankan rupees per kilogram with an average RMSPE of 14.9%. Also, study reveals a correlation between price fluctuation and standard days of the week, where a better prediction was obtained for Monday and Tuesday with an average RMSE of 17.2 and 17.7 Sri Lankan rupees per kilogram respectively with an average RMSPE of 12.2%. Based on the input timestep considered for model, though 14 days and 21 days provided a similar result with minor variation result reveals that 14 days provided a lesser standard deviation of 0.17 than 21 days standard deviation which is 0.98.Publication Embargo A Deep Learning Approach to Outbreak related Tweet Detection(2020 2nd International Conference on Advancements in Computing (ICAC), SLIIT, 2020-12-10) Jayawardhana, B. A. S. S. B.; Rajapakse, R. A. C. P.Due to the popularity of social media around the world, people use to report and discuss real-world events, personal health complications, and disaster situations through these platforms. These social media data streams can be used to track and detect different types of outbreaks. A mechanism is needed to identify outbreak-related tweets to predict the outbreak in advance. In this paper, we propose a deep learning model that can detect tweets related to different outbreaks Epidemics, Public Disorders, and Disasters. GloVe (Global Vectors for Word Representation) embeddings are used as the feature extraction technique as it can capture the semantic meanings of the tweets. Long Short-term Memory (LSTM) which is a specialized Recurrent Neural Network architecture is used as the classification algorithm. In the process, first, outbreak-related tweets were manually collected and curated. Pretrained GloVe word embeddings of 100 dimensions were then used to represent the words of the tweets. As the next step, a Deep Learning Model was trained by using LSTM technique on the curated dataset. Finally, the performance of the model was evaluated using a different dataset. With the results, it can be concluded that the proposed deep learning model is an accurate approach for outbreak-related tweet detection.Publication Embargo PharmaGo-An Online Pharmaceutical Ordering Platform(2021 3rd International Conference on Advancements in Computing (ICAC), SLIIT, 2021-12-09) Gamage, R.G.; Bandara, N.S.; Diyamullage, D.D.; Diyamullage, K.U.; Abeywardena, K.Y.; Amarasena, N.Pharmacy services are a paramount important pillar of health. People must keep social distance due to the COVID-19 pandemic, hence the availability of online services to give medicine is vital. Due to the quarantine measures implemented in and by various countries to prevent the virus’s breaking out and online pharmacies have become an exceptionally popular way to obtain accurate medication. Currently, in Sri Lanka, there are a few mobile applications separately owned by each of the pharmacies to provide online pharmaceutical services for their customers. But all the medicines the customer needs might not be available in a single pharmacy. PharmaGo provides with its cooperation to the customers to get medicines of his necessity at a single pharmacy, as against avoiding him roaming from pharmacy to pharmacy. Similarly, pharmacy owners can read the prescription by using image processing mechanisms and doubtlessly identify the required medicines. In addition, the system analyzes previous sales records and provides predictions regarding the future demand for drugs to the pharmacy owners. PharmaGo includes a highly trained AI-powered medical chatbot to guide the customers throughout the process. PharmaGo provides a reliable platform for both pharmacy users and pharmacists to fulfill the unique needs of pharmacy services.Publication Open Access Agro-Genius: Crop Prediction Using Machine Learning(2019-10) Gamage, A; Kasthurirathna, DThis paper present a way to aid farmers focusing on profitable vegetable cultivation in Sri Lanka. As agriculture creates an economic future for developing countries, the demand of modern technologies in this sector is higher. Key technologies used for this problem are Deep Learning, Machine Learning and Visualization. As the product, an android mobile application is developed. In this application the users should input their location to start the prediction process. Data preprocessing is started when the location is received to the system. The collected dataset divided into 3 parts. 80 percent for training, 10 percent for testing and 10 percent for validation. After that the model is created using LSTM RNN for vegetable prediction and ARIMA for price prediction. Finally, for given location profitable crop and predicted future price of vegetables are shown in the application. Other than the prediction, optimizing for multiple crop sowing according to the user requirements and visualizing cultivation and production data on map and graphs are also given in the application. This paper elaborates the procedure of model development, model training and model testing.