Reinforcement learning and control

Learning to act rather than to predict: building and lighting control, engineering design search, contextual bandits for allocation and advertising, and automated testing. Reward design is usually the hard part.

13 supervised theses. See all supervised theses for the full list, or open topics if you are looking for one to take on.

Level	Author	Title	Date	Industry	Recording
MSc	A. Päkkilä	Active Constraint Set Learning for Constrained Optimization	Aug. 2026	Relex Solutions	video
MSc	L. Veneranta	Optimization of web page advertisements using contextual bandits	Jan. 2024	Sanoma Media
MSc	M. Bogdanova	Contextual bandits for staffing in consulting companies: An exploration of personalized decision making	Oct. 2023
MSc	Y. Sarcheshmehpour	Application of reinforcement learning in electrical machine design	Aug. 2023	ABB	video
MSc	P. Prinsén	Robust gas pressure control using neural networks	Jan. 2022	Wärtsilä
MSc	T. Nguyen	Applying machine learning to develop black-box control model of active double-skin facade	Jan. 2021
MSc	C. Dikmen	Application of contextual bandits models in a supervised learning setting	Aug. 2020
MSc	I. Vikström	Deep reinforcement learning approach for HVAC control	Dec. 2020	TietoEVRY Oyj	video
MSc	J. Eskonen	Deep Reinforcement Learning in automated user interface testing	May 2019	Ericsson
MSc	D. Wu	Unsupervised learning for lighting control system	Oct. 2018	Helvar Oy
BSc	M. Afteniy	Gradient-based deep reinforcement learning	Apr. 2018
MSc	M. Nasir	Supervised learning in lighting control systems	Oct. 2018
MSc	N. Pokhrel	Drone obstacle avoidance and navigation using artificial intelligence	May 2018	Nokia