دانلود کتاب علوم داده و هوش مصنوعی

This volume presents the proceedings of the 3rd International Conference on Data Science and Artiicial Intelligence (DSAI 2025), held from November 19 to 21, 2025, in the historic city of Kandy, Sri Lanka. Following the successful previous editions in Bangkok, Thailand, and Medan, Indonesia, this third edition was jointly organized by the Postgraduate Institute of Science and the Department of Statistics and Computer Science of the University of Peradeniya, and the Department of Statistics of the University of Sri Jayewardenepura. This series continues its mission to foster the growth of data science and artiicial intelligence knowledge, with a special focus on engaging the research communities of developing nations. As AI continues its rapid evolution, reshaping industries and redeining problem-solving paradigms, DSAI 2025 served as a forum for exchanging novel ideas and new methodologies. The conference brought together an international community of researchers and experts to share their work and create new collaborations both globally and at the regional level. The success of this third edition is a witness to the importance and the growing interest in data science and artiicial intelligence. We received 92 submissions from around the world, relecting the conference’s growing stature. Each submission underwent a rigorous single-blind review process, evaluated by at least three members of our international Program Committee. Based on these reviews, 22 high- quality papers were selected for presentation and inclusion in this volume, representing an acceptance rate of 24%. The conference program included four keynote presentations. Mahesan Niranjan explored the transformative role of machine learning in solving complex biomedical problems, while Teeradaj Racharak tackled the critical challenge of building veriiable and trustworthy AI systems. Dilrukshi Gamage provided her view on human-centered design in AI, and Romesh Ranawana discussed the impact of AI beyond its hype. The selected papers cover a diverse range of topics, mirroring the current frontiers of AI research. This volume is structured into sections that include Large Language Models and Advanced NLP, Computer Vision and Deep Learning, Healthcare and Survival Analysis, Time Series Forecasting, Explainable AI and Bias Detection, Optimization and Edge Computing, and Domain-Speciic Applications. These works not only advance theoretical understanding but also demonstrate concrete applications, from detecting pulmonary embolisms and identifying cacao diseases to analyzing FinTech sentiment and forecasting solar irradiance in Sri Lanka’s tropical climate. We are thankful to the authors, speakers, and participants for their contributions to DSAI 2025. Thanks also to the members of the Program Committee and the external reviewers for their efforts and expertise in maintaining the high standard of the proceedings. Our sincere appreciation goes to the organizing universities for their support, Leiden University for the sponsorship, and to all student volunteers whose hard work ensured the smooth execution of the event.