Optimizing Workforce Allocation through Reinforcement Learning: A Machine Learning Approach to Human Resource Management

Efficient management of human resources is crucial for a company’s success, especially in contexts where projects require specific skills and expertise from employees. This thesis aims to address the problem of optimal employee assignment to company projects through the use of Reinforcement Learning (RL) techniques. The main objective is to create an algorithm capable of maximizing the company’s productivity while minimizing assignment execution time and improving overall operational efficiency. The assignment problem has been formalized as an RL environment, where each employee is represented by a set of skills and competence levels, while each project is characterized by specific requirements in terms of skills and experience. Using an agent based on Deep Q-Networks (DQN), the algorithm is able to learn the best assignment strategies by considering the rewards associated with employee performance on each project. The reward is defined based on the completion of projects within the expected timeframe and the quality of work performed, thus contributing to improving overall efficiency. During the experimentation phase, the algorithm showed a progressive improvement in its assignment capabilities. The sum of rewards, both in terms of time steps and absolute value, showed a significant increase, indicating that the agent has learned to optimize resource allocation over time. We also evaluated the algorithm through various performance metrics, such as the average number of correct assignments, the overall time reduction needed to complete projects, and the satisfaction of competence requirements for each project. The results demonstrate that the developed Reinforcement Learning algorithm can significantly improve employee assignment compared to traditional methods, which are based on fixed rules or manual assignments. In conclusion, this work provides a solid foundation for future developments in the field of human resource management assisted by artificial intelligence, with potential applications extending to other industrial sectors where optimal resource management is a critical challenge.