Training Neural Networks Using Reinforcement Learning: The Application of AI in the Twenty-First Century
Overview
With the development of new AI technologies, such as OpenAI’s chatGPT, our lives have become influenced by the factors of AI benefits. In this qualitative study, the researcher explored the field of reinforcement learning by running two simulations with potential solutions to real-world specific problems. T o perform these, the researcher used the program Unity with the plugin UnityML, and also python as a coding language. The results of the simulations provided valuable feedback on the training of neural networks using reinforcement learning. We can see direct correlation between the generated graphs and the learning patterns, such as the cumulative and episode length. In summary, both simulations proved themselves to be potential solutions to these specific issues.
Competition history
- JSHS 2023
Resources
Related projects
ISEF · 2018
Optimizing Reinforcement Learning Through Dynamic Environment Manipulation
CSEF · 2023
Teaching AI to Play Tennis
ISEF · 2023
Optimizing Traffic Flow: Implementing Neural Networks and Deep RL Machine Learning Algorithms to Make Traffic Management More Effectual
CSEF · 2017
Artificial General Intelligence: Efficient Reinforcement Learning to Achieve Human-Like Reasoning
ISEF · 2021
Optimizing Human Feedback in Reward Modelling
ISEF · 2017
Using Reinforcement Learning to Guide Game Rule Design
ISEF · 2019
Novel Reinforcement Learning Methods in Collaborative Environments
ISEF · 2021
Assessing the Impact of Robust Stochastic Operators on Q-learning Efficiency in an OpenAI Gym Environment with a Large Observation Space
Closest projects by meaning, across every fair and year in the corpus.