Machine Learning
null-img

Reinforcement Learning: DQN for temperature control

The goal is to design an intelligent agent that learns to control the shower temperature, adapting to external factors such as the flush of the toilet or similar phenomena. The starting state is initialized with a random temperature ranging from 38 ± 3 °C to introduce variability in the environment.


Used technologies

image - Numpy
image - Gym
image - PyTorch

Reach the project