Improving Location-Aware Applications Through Reinforcement Learning

For my doctoral school course on the priniciples of Artificial Intelligence Problem Solving taught by Hector Geffner, R. Dechter and Andrew Barto, I wrote a paper on the use of Reinforcement Learning techniques to design adaptive location-aware applications:

Improving Location-Aware Applications Through Reinforcement Learning

Abstract. Reinforcement learning (RL) addresses the question of how an autonomous agent that senses and acts in its environment can learn to choose optimal actions to achieve its goals. Each time the agent performs an action in its environment, a trainer may provide a reward or penalty to indicate the desirability of the resulting state. In this paper, I suggest an approach to establish the ability to use RL to construct location-aware systems adapting to user’s expectations in terms of location quality and timeliness.

Relation to my thesis: Machine Learning models could be helpful to design context-aware systems that adapt the location information they deliver to the users (not only adapting the interface) according to the expectations and the environment.