Introduction to Artificial Intelligence & Data Mining

Description

This course is focused on providing students with an introductory overview of theoretical and practical aspects of the intersection of Artificial Intelligence and Data Mining such as: data representation, pattern recognition, data classification, cluster analysis, quantification of classification performance, and data visualization. In addition, the course covers applications of interest such as biomedical data analysis, digital signal processing, image segmentation and video analysis.

Outline

  1. Introduction to artificial intelligence

  2. Paradigms and tasks of machine learning (ML)

  3. Mathematical notation and basics

  4. Time series analysis

  5. Data representation

  6. Performance measures

  7. Clustering

  8. Classification

  9. Data visualization

  10. Algorithmic and implementation issues

  11. Applications


Full syllabus


Recommended textbooks

[1] Ethem Alpaydın. Introduction to Machine Learning. (2nd Edition). MIT Press. Available here.

[2] Richard O. Duda, Peter E. Hart, and David G. Stork. 2000. Pattern Classification (2nd Edition). Wiley-Interscience. First chapter here.

[3] J. Lee, M. Verleysen. Nonlinear Dimensionality Reduction. Springer. here

[4] SDAS Research Group. Introduction to machine learning from linear algebra point of view. (Handout pdf)

[5] Deisenroth, Marc Peter and Faisal, A Aldo and Ong, Cheng Soon. Mathematics for machine learning. 2020 Cambridge University Press. here

[6] KB Petersen, MS Pedersen. The matrix cookbook. Technical Manual, 2012. Technical University of Denmark. here

Resources

Course information (detailed syllabus, evaluation policies, important dates)

General course instructions and guidelines

Full material

  • Lecture 6: Data classification

    • 6.1 Linear classifier

    • 6.2 SVM

      • 6.2.1 Bi-class, hard and soft margin SVM formulation (Slides Handout)

      • 6.2.2 SVM solution via quadractic programming and spectral solution, Multi-class extension

      • 6.2.3 Non-linear formulations (kernel-based)

      • 6.2.4 Applications and extensions: Multiple expert learning, multi-label, semisupervised

  • Additional material