Sökning: "CartPole"
Visar resultat 1 - 5 av 7 uppsatser innehållade ordet CartPole.
1. Fine-tuning Bot Play Styles From Demonstration
Master-uppsats, Uppsala universitet/Institutionen för informationsteknologiSammanfattning : In recent years, Reinforcement Learning (RL) has successfully been used to train agents for games. Nonetheless, in the game industry there is still a necessity for bots not only to succeed in the environments but also to act human-like while playing the game. LÄS MER
2. Playing Atari Breakout Using Deep Reinforcement Learning
Kandidat-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : This report investigates the implementation of a Deep Reinforcement Learning (DRL) algorithm for complex tasks. The complex task chosen was the classic game Breakout, first introduced on the Atari 2600 console.The selected DRL algorithm was Deep Q-Network(DQN) since it is one of the first and most fundamental DRL algorithms. LÄS MER
3. Generation and Detection of Adversarial Attacks for Reinforcement Learning Policies
Kandidat-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : In this project we investigate the susceptibility ofreinforcement rearning (RL) algorithms to adversarial attacks.Adversarial attacks have been proven to be very effective atreducing performance of deep learning classifiers, and recently,have also been shown to reduce performance of RL agents. LÄS MER
4. Control of an Inverted Pendulum Using Reinforcement Learning Methods
Kandidat-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : In this paper the two reinforcement learning algorithmsQ-learning and deep Q-learning (DQN) are used tobalance an inverted pendulum. In order to compare the two, bothalgorithms are optimized to some extent, by evaluating differentvalues for some parameters of the algorithms. LÄS MER
5. Deep Reinforcement Learning in Cart Pole and Pong
Kandidat-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)Sammanfattning : In this project, we aim to reproduce previous resultsachieved with Deep Reinforcement Learning. We present theMarkov Decision Process model as well as the algorithms Q-learning and Deep Q-learning Network (DQN). We implement aDQN agent, first in an environment called CartPole, and later inthe game Pong. LÄS MER