Much of the recent successes within the field of artificial intelligence (AI) in the last 10-15 years can be attributed to a subarea of AI called machine learning. This course is an introduction to machine learning for professionals in the industry and public organisations who have knowledge in engineering, mathematics/statistics, computer science, or related fields.
The course will be held 09.00-15.00 on October 29, November 12, November 26, and December 10 (presentations of homework). Location: Seminar room at MIT-Place, in the MIT building, or through Zoom. Register before October 1. Please register here.
About the course
Machine learning deals with algorithms and methods that learn to perform tasks by using data. The tasks can be anything from detecting an anomalous temperature rise from a sensor, fraudulent money transfers, or email spam to predicting the presence of a disease or to predict which direction to steer a self-driving car. The data can be anything from surveys, polls, and counts to images, text, and sound.
The course gives an introduction to some of the most common and well-established tools and methods in machine learning, and shows the utility of such methods for solving practical problems, and intends to provide the tools needed to use and develop machine learning-based solutions for your own data problems.
Expected learning outcomes
The course will cover many of the main machine learning methods, and the associated theory and algorithms, ranging from classical regression and classification methods to more recent developments in deep learning. The three main learning outcomes for students attending this course are:
Describe and explain a subset of the concepts and methods that are central in machine learning, such as classification, regression, clustering, dimensionality reduction, bias/variance, over- and underfitting, etc.
Categorise machine learning methods based on the needs of the data problem at hand, be able to select an appropriate machine learning method for a task, and properly evaluate a chosen machine learning method. - Apply your knowledge using modern and state-of-the-art machine learning libraries on real data.
The course corresponds to 3 ECTS credits and consists of seven lectures and practical exercises. The course is divided over three days of lectures and exercises, and a fourth day with final lecture and presentations of home assignments. A certificate of attendance will be provided. No formal credits will be possible to obtain in this edition of the course. Note that presence during the course days and submitting the home assignments is mandatory to receive the certificate.
AI Competence for Sweden
Umeå University is one of ten universities that are part of the governmental initiative AI Competence for Sweden, aimed at raising the knowledge about AI in society and build a national platform for education and research.