Sharif Digital Repository

In recent years with advancement of Robotics, the applications of wide scale use of robots in our homes is not as far-fetched as it was before. One of the important problems for an indoor robot is moving inside a new and unknown environment. To achieve this, the robot should, with the help of its sensors, not only calculate its location; but should also build a map of the environment for later use. Additionally, the robot must be able to explore this unknown environment. The most important drawbacks of the classical solutions to these problems are long computation times, heavy memory usage and absence of precision. In recent years, large amount of research effort has put on solving these... 

Reinforcement learning is one of the most well-known learning paradigms in biological agents and one of the most used ones for solving plenty of problems. One of the reasons for this widespread use is the low demand for supervising signals. However, the sparsity of the reward signal causes increasing in sample complexity that needs for learning new tasks. This issue makes trouble in multi-task settings, specifically.One of the most promising approaches to learning new tasks by limited interaction with the environment is meta reinforcement learning. An approach in which fast adaption becomes possible by limiting hypothesis space and creating inductive biases by learning meta parameters....