work
Environments
category | Video |
subject | Human figure, Architecture, Animal |
tags | AI, Artificial Intelligence, Reinforcement Learning, Intelligenza Artificiale, IA, experimental |
minutes | 10 |
seconds | 0 |
year | 2021 |
Video of Reinforcement Learning Training.
Digital Video, NFT 2Kpx Square.
Environments is a series of works focused on the visualization of automatic training of neural networks through reinforcement learning techniques.
Reinforcement learning is one of the three most important machine learning paradigms along with supervised and unsupervised learning. This technique allows artificial intelligences to learn to perform actions that lead to achieving a goal by acting as effectively as possible within a given environment.
In Environments each work shows a small world within which the intelligent agent can operate without conditioning. The AI does not have a prior knowledge of the rules and what surrounds it, but after several attempts to get the correct feedback from the environment it will learn on its own how to move as it suits her best to get the reward.
Reinforcement learning is often modeled through Markov Decision Making (MDP), particularly useful for addressing a wide range of optimization problems in contexts where results are partly random and partly under the control of a decision maker.
Digital Video, NFT 2Kpx Square.
Environments is a series of works focused on the visualization of automatic training of neural networks through reinforcement learning techniques.
Reinforcement learning is one of the three most important machine learning paradigms along with supervised and unsupervised learning. This technique allows artificial intelligences to learn to perform actions that lead to achieving a goal by acting as effectively as possible within a given environment.
In Environments each work shows a small world within which the intelligent agent can operate without conditioning. The AI does not have a prior knowledge of the rules and what surrounds it, but after several attempts to get the correct feedback from the environment it will learn on its own how to move as it suits her best to get the reward.
Reinforcement learning is often modeled through Markov Decision Making (MDP), particularly useful for addressing a wide range of optimization problems in contexts where results are partly random and partly under the control of a decision maker.