The aim of this project is to solve problems involving sequential decision-making under uncertainty by building agents that learn by reinforcement. Problems of this nature involve situations in which an agent (be it a human, an organization or a computer program) needs to make a series of decisions over time, and each decision can influence the future options available and the possible outcomes.
APPLICATION EXAMPLES:
- Financial investment: an investor who needs to decide which stock to buy, when to buy it, sell it, and so on;
- Route planning: an autonomous vehicle must decide where to move at any given moment, and each movement can open up or close off different future paths;
- Logistics: design of a decision-making agent for inventory control in a retail environment, where product demand is uncertain and the agent needs to optimize its inventory replenishment decisions over time.
MANDATORY INFRASTRUCTURE RESOURCES:
- Cloud environment and services;
- Deep Racer from AWS;
- Python language;
- Data visualization with a graphic library suitable for the project;
- Version control system: GitHub platform.
COMPLETION AND DELIVERY OF THE PROJECT:
All the prototypes generated during the project are delivered at the end of the 10th week.
