Courses Taught

This course delves into the use of prominent machine learning (ML) models through Python, specifically targeting health data applications and extending beyond. The curriculum covers: Foundations of Python programming, including pivotal modules like NumPy, Pandas, Matplotlib, and Scikit-learn; Classification-focused ML models; regression-centric ML models; Techniques for ML model training and validation; Support vector machines and decision trees; Ensemble methodologies; Strategies for dimensionality reduction; and Unsupervised learning techniques. By the course’s conclusion, students will: Grasp the underlying mathematical and statistical algorithms, along with the computational routines, of prevalent ML models in the health and social sciences; Apply ML models adeptly to dissect real-world data; and Evaluate the advantages and limitations of various ML models in diverse problem-solving scenarios. Furthermore, the course establishes a robust foundation for those eager to delve into deep learning, a niche within ML centered on artificial neural networks and representation learning.

This course provides comprehensive instruction on a diverse range of deep learning (DL) models, utilizing Python and focusing on applications in health data and beyond. The curriculum encompasses: An introduction to deep learning, Python, and NumPy; The basics of PyTorch and neural networks; Computer vision, covering image classification, object detection, image segmentation, keypoint detection, audio classification, and video classification; Natural language processing, focusing on text preprocessing, text classification, text generation, text summarization, and text question answering; Techniques for time series forecasting; Development of recommender systems; Exploration of generative adversarial networks; and Methods for synthetic data generation. Upon completing this course, students will: Acquire an in-depth understanding of the fundamental concepts and components of AI, ML, and DL; Become well-acquainted with a wide array of advanced, state-of-the-art DL models and their applications in health and other domains; Gain insight into the strengths, limitations, and trade-offs of varying DL models, along with the best practices for implementing them; and Implement DL models, such as convolutional neural networks, recurrent neural networks, and transformers, on different data types including text, image, video, audio, and tabular, using Python in conjunction with popular APIs and cloud platforms like PyTorch, PyTorch Lightning/Flash, fastai, Hugging Face, spaCy, Haystack, Synthetic Data Vault, Google Colab, and Kaggle.

A systematic review attempts to identify, appraise, and synthesize all the empirical evidence that meets pre-specified eligibility criteria to answer a specific research question. Meta-analysis is a quantitative, formal, epidemiological study design used to systematically assess the results of previous research to derive conclusions about that body of research. This course seeks to: Enhance students’ comprehension of key concepts and techniques related to systematic review and meta-analysis through focused lectures and assignments; and Equip students with practical skills necessary for conducting systematic reviews and meta-analyses through hands-on practices and programming exercises. Topics covered include: Fundamental concepts of systematic review and meta-analysis; Formulation of review topics; Development of review protocols; Determination of inclusion and exclusion criteria; Development of keyword search algorithms; Execution of forward and backward search methods; Management of references using Endnote; Utilization of data extraction tools; Assessment of study quality and bias; Determination and pooling of effect sizes across studies; Introduction to conducting meta-analyses using R, encompassing fixed-effect and random-effect meta-analysis; Exploration of meta-analysis with binary outcomes; Analysis of heterogeneity and meta-regression; Addressing small-study effects; Handling of missing data in meta-analysis; Implementation of multivariate meta-analysis; Meta-analysis of diagnostic test accuracy studies; and Execution of network meta-analysis. By the end of this course, students will possess an in-depth understanding of systematic review and meta-analysis concepts and techniques and will be adept at conducting such reviews and analyses using R.

This course introduces fundamental concepts and methods of biostatistics to students majoring in social and health sciences. Topics include: Introduction to tabular data and R programming; Descriptive statistics; Measures of disease risk; Probability and diagnostic testing; Probability distributions; Statistical inference; Statistical tests; and Linear regression.

This course introduces data analysis methods commonly applied in epidemiology and health policy research using Stata and R. Topics include: Introduction to health data; Confounding bias; Research designs; Descriptive statistics; Incidence, cumulative incidence, and prevalence; Principles of probability; Probability distributions; Confidence interval; Hypothesis testing; Statistical testing for continuous outcomes; Statistical testing for categorical outcomes; Simple linear regression; Multiple linear regression; Model diagnosis; and Model selection.