Course 2021-2022 a.y.

30412 - MACHINE LEARNING

Department of Decision Sciences

Course taught in English
Go to class group/s: 25
BEMACS (6 credits - II sem. - OB  |  INF/01)
Course Director:
CHRISTOPH JOHANN FEINAUER

Classes: 25 (II sem.)
Instructors:
Class 25: CHRISTOPH JOHANN FEINAUER


Suggested background knowledge

Being familiar with Python programming, elementary calculus and basic statistics help students understand most of the covered topics.

Mission & Content Summary

MISSION

Scope of the course is to provide an introduction to the fundamental concepts and tools of modern machine learning techniques. These tools are at the root of data science and data analytics, which are among the main pillars of the education program.

CONTENT SUMMARY

  • Introduction to the theory of Machine Learning.
  • Review of probability tools.
  • Statistical inference and regression techniques.
  • Unsupervised methods: Principal Component Analysis, hierarchical clustering, k-means.
  • Supervised methods: K-nearest neighbours,  Support Vector Machines, Multi Layer Neural Networks.
  • Implementing Machine Learning Pipelines with sklearn

Intended Learning Outcomes (ILO)

KNOWLEDGE AND UNDERSTANDING

At the end of the course student will be able to...
  • State and use the basic results of statistical inference for data analysis.
  • Distinguish between supervised and unsupervised learning methods.
  • Describe the  basic conceptual ideas, strengths and the limitations of the different learaning algorithms.

 

APPLYING KNOWLEDGE AND UNDERSTANDING

At the end of the course student will be able to...
  • Implement a complete data analysis pipeline, from unsupervised clustering of data to supervised classification using the Python scikit-learn library.
  • Demostrate the ability to cope with real-world data analysis and problem solving (managing, preprocessing and analyze real datasets).

Teaching methods

  • Face-to-face lectures
  • Individual assignments

DETAILS

  • Individual assignemnts: each student is required to solve a machine learning problem and provide a written report and a code.

Assessment methods

  Continuous assessment Partial exams General exam
  • Written individual exam (traditional/online)
    x
  • Individual assignment (report, exercise, presentation, project work etc.)
    x

ATTENDING AND NOT ATTENDING STUDENTS

The exam consists of a theory part and problem solving projects.

  • The theory part consists in exercises and questions to be answered on paper, and is used to asses the "knowledge and understanding" learning   objectives. This contributes to 50% of the final grade.
  • The project consists in a programming code to solve a concrete ML problem, to be developed individually and described through a written report, which is evaluated by the teachers. This contributes to 50% of the final grade.

The individual project is used to asses the "applying knowledge and understanding" learning objectives. In order to pass the exam, students must achieve a passing grade  in both the theory part and the project part.


Teaching materials


ATTENDING AND NOT ATTENDING STUDENTS

  • D. MACKAY, Information Theory, Inference, and Learning Algorithms, Cambridge University Press, 2013 (but available for download as decided by the Author).
  • C.M. BISHOP, Pattern Recognition and Machine Learning, Springer, 2006.
  • T. HASTIE, R. TIBSHIRANI, J. FRIEDMAN, The Elements of Statistical Learning, Springer, 2009. 
  • I. GOODFELLOW, Y. BENGIO, A. COURVILLE,  Deep Learning, MIT Press, 2016 (optional).
  • Handouts of each lecture and sample codes are provided.
Last change 21/12/2021 10:29