Publications
Research is about pushing boundaries, questioning assumptions, and contributing knowledge that moves the field forward.
Filter by Category:
Abstract: This research addresses the critical challenge of data privacy in healthcare while maintaining the effectiveness of machine learning models. The study presents an innovative approach combining federated learning with random forest algorithms specifically designed for heart disease detection.
Abstract: This paper deals with the reliability worth assessment of electric power distribution system considering momentary interruptions. Sequential Monte Carlo simulation technique is applied for the assessment purpose considering uncertainty of associated random variables.
Abstract: AgriBuddy is an AI-powered, agent-based system that delivers localized agricultural recommendations in natural Bangla dialogue. The system combines Retrieval-Augmented Generation with specialized agents and CNN-based vision models for rice disease detection.
Abstract: This work examines Bengali phonology and phonetics, presents a comprehensive study of Bengali IPA transcription, and introduces a novel IPA transcription framework with deep learning-based benchmarks for linguistic analysis and NLP resource creation.
Abstract: This thesis investigates regional Bengali dialects and develops Automatic Speech Recognition systems tailored to regional Bengali languages using character-gram modeling approaches. The work contributes to the understanding and preservation of regional Bengali dialects.
Abstract: This thesis presents an efficient method for reliability evaluation of power distribution systems considering momentary interruptions, utilizing sequential Monte Carlo simulation to analyze customer interruption costs and system performance.