http://dlib.iit.ac.lk:80/xmlui The DSpace digital repository system captures, stores, indexes, preserves, and distributes digital research material. 2026-10-07T18:56:14Z https://dlib.iit.ac.lk/xmlui/handle/123456789/3745 Customer Churn Prediction System: With Special Reference to Telecommunication Industry in Sri Lanka Sivapragasam, Savilachine The Sri Lankan telecommunications industry has emerged as an important problem of frequent customer churn, which has direct effects on the telecommunications industry’s revenue and profitability, or its future sustainability. Customers also seem to be churning at a fast pace in a competitive market making it very important for service providers to be on the lookout for anytime churn customers and determine why they are churning. The main goals of this research are the construction of a model of analyzing diverse factors that may lead to customer dissatisfaction and an evaluation of potential churn, with the primary focus on allowing telecom operators to intervene to address customers’ concerns. analyses are performed, and each telecom company can use the system to identify likely churners along with recommendations for timely retention efforts to improve customer satisfaction. 2025-01-01T00:00:00Z https://dlib.iit.ac.lk/xmlui/handle/123456789/3744 Voluntra - Streamlining Volunteer Efforts Using AI for University Students for Sustainable Social Impact. Rajapaksha, Ruwini Voluntra is a mobile app which seeks to increase volunteer activity among Sri Lankan university students by bridging some of the primary challenges such as low exposure to quality-checked opportunities, mismatch between skills and tasks, poor appreciation, and poor availability of good flexibility in scheduling. Most universities lack an organized volunteering mechanism, which means students depend on unorganized, ad-hoc sources of finding appropriate opportunities. While universities are struggling to manage volunteer schemes, including finding activities, sorting through student applications manually, and tracking involvement, Voluntra offers a simple-to-use, skill-matched platform where students can apply to projects thoroughly vetted by their university admin, track progress, and earn badges for involvement. Voluntra has AI-based app screening, a badge system with tiers, and quality assurance through feedback. Admins can create projects, approve students, and monitor activity in a secure dashboard. Voluntra bridges the difference between student engagement and institutional capacity by making volunteering more accessible, better organized, and rewarding—enabling long-term civic engagement across universities. 2025-01-01T00:00:00Z https://dlib.iit.ac.lk/xmlui/handle/123456789/3743 Load Forecasting for National Grid Using Recurrent Neural Network Jayalath, Chathurya Accurate short-term load prediction is crucial for effective energy management in order to ensure proper operation of national grids and optimal use of available resources. This research explores the application of Long Short-Term Memory (LSTM) networks, a Recurrent Neural Network (RNN), to improve the accuracy of short-term electricity load forecasting in Sri Lanka. Traditional methods, such as Multiple Linear Regression (MLR), have been found inadequate to handle the non-linear, complex associations of electricity demand. The LSTM model, able to dig up temporal relations in time-series information, is trained on historical load demand data set extracted by Kaggle and other environmental and economic variables such as temperature, humidity, and GDP. The model is better in accuracy than the MLR approach, and performance metrics such as Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE) have a significant improvement. The research proves the potential of LSTM-based models for real-time load prediction, which can enhance grid stability, reduce resource wastage, and enhance the efficiency of operation. The findings are advantageous to the increased application of deep learning models within energy systems, particularly within the developing world, to offer valuable insights to future systems for energy forecasting and management. 2025-01-01T00:00:00Z https://dlib.iit.ac.lk/xmlui/handle/123456789/3742 A data-driven inventory optimization strategy: Coconut Oil Sales Quantities Forecast Fernando, Vimanya A leading entity in the food manufacturing industry in Sri Lanka, encounters substantial challenges in inventory management and forecasting. The reliance on basic internal forecasting methods, due to the high cost of Salesforce’s forecasting extension, has led to inaccuracies in inventory forecasting. These inaccuracies result in oversupply, which ties up capital and increases storage costs, and undersupply, which causes stockouts, lost sales, and disrupted cash flow. These issues impact the financial department, inventory management team, and overall business operations, especially during seasonal periods. The aim of this project was to develop a more accurate and efficient forecasting model to improve inventory management, reduce financial strain, and enhance operational efficiency at this Company. The methodology involved incorporating external factors into a time series model, including Sri Lankan coconut prices, coconut oil selling prices and the Colombo Consumer Price Index. Both a basic LSTM model and an SARIMAX model with feature engineering were developed to determine the more accurate model. The forecasting models were implemented using Python and trained on pre-processed historic data to predict coconut sales quantities for 3 months. The performance of both models was assessed using evaluation metrics, and the LSTM model was selected as the more accurate model, with a Mean Absolute Error (MAE) of 18.70, a Root Mean Squared Error (RMSE) of 23.99, and an R² Score of 0.56. The final model output was integrated into a Microsoft Power BI dashboard, providing insights to support production management decision-making and detailing sales performance. The final results indicate that incorporating a systematic forecasting model with external influencing factors outperforms traditional methods. The conclusions drawn from this project suggest that advanced forecasting models can effectively address inventory management challenges, reduce financial strain, and enhance overall operational efficiency at this Company. These observations highlight the importance of accurate sales forecasting in optimizing inventory levels and supporting strategic business decisions. 2025-01-01T00:00:00Z