Study Guide

Objective

Machine learning methods have become an important tool for automation, control and data-driven decision making in an industrial context. The course aims at introducing and practicing several of these tools for system identification, classification and predictive analytics.

Knowledge and understanding
After completing the course, the student should be able to:
- identify and recognize industrial problems that are solvable through machine learning
- plan and apply the machine learning workflow to classification, regression and identification problems
- understand and explain the difference between different machine learning tasks and techniques

Skills and abilities
After completing the course, the student should be able to:
- develop and build machine learning models for different tasks
- validate and test the performance of a model
- develop hands-on experience with machine learning techniques

Evaluation ability and approach
After completing the course, the student should be able to:
- interpret the result of a machine learning experiment
- report on the result of a machine learning experiment
- assess the outcome of a machine learning experiment

Content

The availability of big amounts of data for various systems makes it possible to apply both linear and nonlinear models to fit the data. The course focuses on different techniques that can be used for identification, classification and predictive purposes in the area of automation and control.

Materials

See moodle page

Teaching methods

Lectures, exercises, laboratory work

Student workload

A total of 125 h

Evaluation scale

H-5

Assessment criteria, satisfactory (1)

Satisfactory skills in understanding and using machine learning tools for classification, prediction and identification.

Assessment criteria, good (3)

Good skills in understanding and using machine learning tools for classification, prediction and identification.

Assessment criteria, excellent (5)

Excellent skills in understanding and using machine learning tools for classification, prediction and identification.

Assessment methods and criteria

See moodle page or course info on peppi

Qualifications

No prerequisites