course aims in Estonian
Õppeaine eesmärk on:
- tutvustada sardsüsteemide jaoks sobilikke masinõppe mudeleid;
- tutvustada klassifitseerimisalgoritme ja regressioonalgoritme;
- tutvustada masinõppes kasutatavaid sardsüsteeme;
- omandada masinõppemudelite optimeerimise põhitõdesid;
- tutvustada masinõppe jaoks kasutatavaid töövahendeid;
- tutvustada treenimisandmete loomist ja kasutamist sardsüsteemide põhistes lahendustes.
course aims in English
The aim of this course is to:
- introduce machine learning models for embedded systems;
- introduce classification and regression algorithms;
- introduce embedded systems used for machine learning;
- acquire the basics of model optimization for embedded machine learning;
- introduce and use tools for machine learning;
- introduce the creation and use of training data in solutions based on embedded systems.
learning outcomes in the course in Est.
Õppeaine edukalt läbinud üliõpilane:
- omab ülevaadet erinevates sardsüsteemide jaoks sobilikest masinõppemudelitest;
- valib konkreetse ülesande lahendamiseks sobiva masinõppemudeli ja kasutab seda;
- valib sobiva riistvara masinõppe mudeli realiseerimiseks;
- oskab kasutada töövahendeid mudelite treenimiseks ja ehitamiseks;
- loob ja kasutab õigesti treenimisandmeid;
- optimeerib mudeleid sõltuvalt kasutatava riistvara piirangutest.
learning outcomes in the course in Eng.
After successfully completing the course, the student:
- has an overview of various machine learning models suitable for embedded systems;
- chooses and uses appropriate machine learning model to solve a specific task;
- chooses appropriate hardware to implement machine learning model;
- uses tools to train and build models;
- creates and uses training data correctly;
- optimizes models depending on hardware limitations.
brief description of the course in Estonian
Masinõppes peamiselt kasutatavate algoritmide ülevaade ning nende tööpõhimõtetest arusaamine. Masinõppe mudelite loomise alused. Treenimisandmete loomine ja nende kasutamine. Olulisemad mudelite optimeerimise põhitõed ja nende rakendamine sardsüsteemide jaoks. Spetsiaalset riistvara ehitama ei pea. Kursuse jooksul lahendatakse individuaalsed ülesanded ning 2-3 liikmelised grupid projektülesande.
brief description of the course in English
Overview of main algorithms used in machine learning. Basics of model creation. Building of training data sets and using them. Main concepts of model optimization and their use for embedded systems. No special hardware is developed. Individual tasks and project assignments (by teams of 2-3 members) are solved during the course.
type of assessment in Estonian
-
type of assessment in English
-
independent study in Estonian
Iseseisev töö seisneb teoreetiliste materjalide läbitöötamises, praktikumides ette antud iseseisvate ülesannete täitmises ja projektülesande täitmises.
independent study in English
Independent work consists of working through theoretical materials, performing independent assignments given in practical classes and completing a project assignment.
study literature
TinyML: Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers, Pete Warden, D. Situnayake.
Embedded Deep Learning, B. Moons, D. Bankman, M. Verhelst.
Embedded Vision: An Introduction, S. R. Vijayalakshmi, S. Muruganand.
study forms and load
daytime study: weekly hours
4.0
session-based study work load (in a semester):