Portfolio Optimization of Renewable Energy Generation for Economic Sustainability

This research aims to demonstrate how data science can be an effective tool to assist in corporate decision-making, provided it is applied in a structured and efficient manner to collect, process, and transform data into performance metrics and indicators that can guide the decision-making process, adhering to the complexity of organizational structures and the large volume of generated data. This research presents a solution to systematize the critical decisions of a company that operates in the renewable energy generation sector, using the CRISP-DM methodology, which is the most used by the market in data science projects, in addition to the Markowitz portfolio theory, statistical methods, and nonlinear mathematical programming. The challenge addressed was to direct the expansion of the company’s generator park toward an optimized diversification of generation sources that ensures consistent financial performance, achieving the best-expected outcome versus volatility ratio.