A personal review of Georgia Tech's OMSCS CS6601 Artificial Intelligence course, covering course structure, assignments, and exams. The course is rated highly difficult (4.3/5) with an average workload of 23 hours per week. It covers fundamental AI topics including search, adversarial search, Bayes networks, HMMs, constraint satisfaction, and decision trees — areas often skipped by popular online courses. Assignments require implementing algorithms from scratch in Python, and exams are lengthy take-home papers. The author shares practical tips: start assignments on the first weekend, attend TA office hours, and always include methodology for partial credit. A standout highlight was placing second in a Kaggle competition using a self-implemented decision tree classifier enhanced with techniques from XGBoost and LightGBM papers.
