IDTS4101 - Data Classification

This course offers a rigorous, advanced treatment of supervised machine learning and data mining, focusing strictly on the algorithms, mathematical foundations, and specialized domains of data classification. The curriculum is closely aligned with the definitive text Data Classification: Algorithms and Applications by Charu C. Aggarwal.

Students will explore the core mathematical mechanics of traditional classifiers alongside highly specialized variations. The course progresses from fundamental decision trees and instance-based lazy learners to rule-based classifiers, probabilistic models (Bayesian networks), support vector machines (SVM), neural architectures, and ensemble configurations (Bagging, Boosting, and Random Forests). Beyond basic flat text or tabular arrays, students will master specialized classification pipelines engineered for complex, high-dimensional data types, including discrete text corpuses, streaming real-time data, network graphs, spatial-temporal matrices, and multimedia assets.

The course also explores critical algorithmic governance parameters, highlighting performance evaluation boundaries (ROC/AUC analysis, rare-class imbalances, cost-sensitive matrices), feature selection techniques, profiling bias auditing, and data privacy masking compliance under regulatory frameworks like the Indonesian UU PDP and GDPR. Through hands-on programming labs, students will develop the professional capacity to design, optimize, and defend scalable classification solutions for complex corporate and research applications.