Mathematical Principles of AILaajuus (5 cr)
Course unit code: DIG4HM111
General information
- ECTS credits
- 5 cr
- Teaching language
- English
Learning objectives
The participant will learn the mathematical methods behind most common machine learning models and learn how to apply them in the development of AI-based applications. He/she will know several optimisation methods to find the optimal solution and will be able to justify the choice in a given situation. After completing the course, the participant will be able to work independently in the development of models for AI applications and apply the available knowledge in model building.
- master the mathematical methods underlying machine learning models
- be able to develop Python-based solution models for optimisation problems in machine learning
- be able to apply data collected from information systems to model building
- understand the role of mathematical models in different parts of neural networks
Contents
Mathematical models for applied artificial intelligence
- One and several variable differential and integral calculus
- Basics of linear algebra: vectors, matrices
- Regression models
- Different numerical methods for optimisation
- Symbolic and numerical calculations based on Python programming language
- Neural networks
Lähtötaso ja sidonnaisuudet muihin opintojaksoihin
This course unit is part of the master's degree's curriculum. Completion of the course requires master's study entitlement.
Implementation methods, demonstration and Work&Study
This course is organised as a CONTACT implementation type. Attendance in classes and exam 48 h, independent study 87 h. Total 135 hours.
The course includes Moodle quizzes and homework assignments, each of which must be returned in accordance with the timetable for successful completion. Finally, the exam must also be passed in the same implementation.
If one already knows the learning outcomes of this course, the skills according to the knowledge can be demonstrated. Ask your course teacher for more information. For more information on the recognition of competences: https://www.haaga-helia.fi/fi/osaamisen-tunnistaminen-ja-work-study
Assessment criteria - grade 1
Student is be able to perceive the key concepts of machine learning mathematics. He/she will be able to perform individual tasks related to the calculation of key concepts. Student can also apply their learning in practice to some extent.
When the implementation type of the course is CONTACT, ONLINE or BLENDED it is required that the student is present during those teaching hours that are marked in the study schedule. If you are absent more than 25 %, your grade will be lowered by one. If you are absent more than 50 %, the course is failed.
Assessment criteria - grade 3
Student can model computations in machine learning problems at a basic conceptual level. He/she is able to carry out planning and controlling calculations to develop the models and to calculate related assessment calculations. Student can also apply calculations to practical situations.
Assessment criteria - grade 5
Student has a deep understanding of the mathematical concepts of machine learning models and he is able to seamlessly perform calculations for AI applications. He/she is able to extensively apply the learned skills to problems in machine learning and to independently perform related calculations. Student is also able to interpret the solutions of the calculations for interest groups. The student will also be very good at applying his/her learning to practical computational situations.
Learning materials
Lectures and other material to be shared on the Moodle learning platform.
Additional literature:
Deisenroth, Marc Peter, Faisal, A. Aldo, and Ong,?Chen, Mathematics of Machine Learning, Cambridge University Press,?2020