Automatic Tuning of Q-learning Algorithms Parameters

A Q-learning Based Continuous Tuning of Fuzzy Wall Tracking

A simple easy to implement algorithm is proposed to address wall tracking task of an autonomous robot. The robot should navigate in unknown environments, find the nearest wall, and track it solely based on locally sensed data. The proposed method benefits from coupling fuzzy logic and Q-learning to meet requirements of autonomous navigations. Fuzzy if-then rules provide a reliable decision maki...

Automatic selection of tuning parameters in wind power prediction

Machine Learning for Automatic Physical DBMS Tuning

Tuning a DBMS that experiences varying workload is challenging. Database administrators cannot be expected to monitor the workload and react with appropriate tunings, therefore automation is essential. In this report we outline a new method for automatic physical DBMS tuning that uses machine learning to model and predict workloads, and tune for the future. Our method builds on previous approac...

Automatic Tuning of Control Parameters for Single Speed Engines

In Scania’s single speed engines for industrial and marine use, the engine speed is controlled by a PI-controller. This controller is tuned independent of engine type and application. This brings certain disadvantages since the engines are used in a wide range of applications where the dynamics may differ. In this thesis, the possibility to tune the controller automatically for a specific engin...

New algorithms of the Q-learning type

We propose two algorithms for Q-learning that use the two timescale stochastic approximation methodology. The first of these updates Q-values of all feasible state-action pairs at each instant while the second updates Q-values of states with actions chosen according to the ‘current’ randomized policy updates. A proof of convergence of the algorithms is shown. Finally, numerical experiments usin...