Khaing Phyo Wai

I hold a Bachelor’s degree in Computer Science, followed by an Honours degree in same discipline that provided advanced theoretical grounding. I further pursued a Master’s degree in Computer Science, deepening my expertise in computational methodologies and research practices. I completed my academic training with a PhD in Information Technology, where I engaged in extensive research and contributed to the advancement of knowledge in the field. All of these degrees were obtained from the University of Computer Studies, Mandalay (UCSM), Myanmar, reflecting a continuous and coherent academic progression within a consistent scholarly environment.

My research interests lie broadly in artificial intelligence, machine learning, and data-driven computational methods. I am particularly interested in data science and data-driven approaches for processing and analyzing different types of data, with a focus on extracting meaningful patterns and supporting evidence-based insights. I am also interested in the development of interpretable and explainable models, intelligent decision-making systems, and analytical frameworks that enhance the transparency and reliability of AI. Overall, my work explores how advanced computational techniques can be applied to complex datasets to improve predictive performance, strengthen interpretability, and enable responsible, human-centered applications of AI across diverse domains.

During my PhD’s studies, I investigated distance-based clustering of moving-object trajectories using large-scale trajectory datasets, examining how similarity measures and spatio-temporal characteristics can be used to uncover meaningful mobility patterns. This work strengthened my understanding of high-dimensional data analysis and the development of clustering methodologies capable of handling complex and voluminous movement data. Building on this foundation, my postdoctoral research focused on explainable artificial intelligence (XAI), with an emphasis on developing interpretable models and analytical frameworks that provide transparent insights into Multi-Agent Deep Reinforcement Learning Models. This experience expanded my interest in model interpretability, intelligent decision-making processes, and the integration of statistical learning with explainable computational techniques. Currently, I am working as a Data Scientist at the School of Economic, Singapore Management University (SMU), where I apply data-driven approaches and machine learning methods to process and analyze diverse datasets, and derive actionable insights for the economic doamin.

I have gained extensive teaching experience through my involvement in both undergraduate and postgraduate instruction across a range of computing disciplines. I have taught Information Technology Passport (IP) and Fundamental Information Technology Engineer (FE) courses, supporting students preparing for the professional IT proficiency examinations administered by the Information-Technology Promotion Agency (IPA), Japan. My teaching portfolio also includes core computing courses such as Data Structures and Algorithms, Software Engineering, Database Systems, Information Processing Systems, Information System Strategy, and Security. As part of an established teaching team, I have additionally contributed to the delivery of advanced modules including Multi-Agent Systems, Data Mining, and Machine Learning. My responsibilities have encompassed developing teaching materials, preparing and evaluating assessments, addressing students’ learning challenges, providing guidance on study skills, offering constructive feedback, and conducting mock tests to support students’ academic and professional development.

I have executed a broad range of self-driven machine learning projects across critical domains, including finance, healthcare, and engineering. My work demonstrates expertise in applying and comparing various Supervised Learning models while maintaining an emphasis on Explainable AI (XAI) techniques to ensure full model transparency and interpretability.

• Finance & Economics
  • Credit risk prediction using a Gradient Boosting Classifier trained on the German Credit dataset. SHAP (Shapley Additive explanations) was used for feature importance analysis and model transparency.
  • Income prediction system developed with Neural Networks. LIME (Local Interpretable Model-agnostic Explanations) was employed to provide local explanations and enhance interpretability.
• Healthcare
  • Pneumonia detection in chest X-ray images using a Convolutional Neural Network (CNN), resulting in improved diagnostic accuracy.
  • Early cancer detection using the Breast Cancer Wisconsin (Diagnostic) dataset, applying a Random Forest classifier optimized via Grid Search.
  • Diabetes onset prediction for Pima Indian women using SVM, Logistic Regression, and XGBoost to facilitate a comparative analysis of classifier performance across different algorithms.
  • Heart disease prediction achieved through a deep learning model implementation, utilizing Captum’s DeepLift to provide feature-level interpretability.
• Science & Engineering
  • Predictive model for battery performance constructed using a synthetic dataset. Random Forest regression was applied, and LIME (Local Interpretable Model-agnostic Explanations) was used to interpret model behavior.
  • Iris species classification implemented using a Random Forest classifier, incorporating SHAP (Shapley Additive explanations) for explainability and adversarial training to enhance model robustness.
• Fairness, Safety & Trustworthy AI
  • Bias-mitigation study conducted to support the development of safe, trustworthy, and fair AI systems in high-stakes contexts.
  • Fairness and algorithmic bias explored in criminal recidivism prediction using the COMPAS dataset, assessing model behavior across demographic groups.
  • Hybrid models constructed by combining Logistic Regression, Random Forest, and SHAP-based explainability to assess recidivism risk in a transparent and interpretable manner.

I have participated in several team-based applied projects that span a broad range of technical domains. In one project, I contributed to the development of a quick chat web application featuring real-time communication capabilities. I also collaborated on the design of a computerized system for record-keeping and textbook sales to enhance data management and streamline operational processes. As part of another team project, I worked on implementing a computerized system for driving license registration and processing. Additionally, I contributed to a networking-focused project involving IPv6 configuration and routing. Beyond these collaborative efforts, I independently carried out a Lean Six Sigma Green Belt project aimed at improving the prediction accuracy of a classification process through structured analytical and process-improvement methodologies. Together, these projects reflect my ability to contribute effectively in both individual and team settings while applying computational and problem-solving skills across diverse practical contexts.