English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Learning optimal policies from sparse feedback is a known challenge inreinforcement learning. Hindsight Experience Replay (HER) is a multi-goalreinforcement learning algorithm that comes to solve such tasks. The algorithmtreats every failure as a success for an alternative (virtual) goal that hasbeen achieved in the episode and then generalizes from that virtual goal toreal goals. HER has known flaws and is limited to relatively simple tasks. Inthis thesis, we present three algorithms based on the existing HER algorithmthat improves its performances. First, we prioritize virtual goals from whichthe agent will learn more valuable information. We call this property the\textit{instructiveness} of the virtual goal and define it by a heuristicmeasure, which expresses how well the agent will be able to generalize fromthat virtual goal to actual goals. Secondly, we designed a filtering processthat detects and removes misleading samples that may induce bias throughout thelearning process. Lastly, we enable the learning of complex, sequential, tasksusing a form of curriculum learning combined with HER. We call this algorithm\textit{Curriculum HER}. To test our algorithms, we built three challengingmanipulation environments with sparse reward functions. Each environment hasthree levels of complexity. Our empirical results show vast improvement in thefinal success rate and sample efficiency when compared to the original HERalgorithm.

Deep Reinforcement Learning for Complex Manipulation Tasks with Sparse Feedback