In this work, after constructing a dataset composed of thirty-year daily time series of the major commodities traded on financial markets and some of the key assets in the U.S. stock, currency, and bond markets, machine learning models will be trained with the goal of classifying economic scenarios to apply the most suitable trading operational strategy for the identified scenario.
Additionally, given the critical nature of the financial context in which decisions are made, Explainable Artificial Intelligence (XAI) algorithms will be employed to make these decisions transparent and interpretable.
Project Phases:
- Data Set Construction: Collect and organize daily 30-year data on major U.S. commodities, equity, currency and bond assets. Ensure that data are complete, accurate, and ready for analysis.
- Data Preprocessing: Normalize and standardize data to ensure consistency during model training. Address any missing data or outliers that could adversely affect model performance.
- Training Machine Learning Models: Use machine learning algorithms, such as neural networks, decision trees or support vector machines, to classify economic scenarios based on time series. Partition the data into training and test sets to evaluate the effectiveness of the models.
- Implementation of XAI Algorithms: Integrate Explainable Artificial Intelligence algorithms to make model decisions transparent and interpretable. Analyze and visualize the most influential features that lead to a given classification.
- Validation and Optimization: Verify the accuracy of models using appropriate metrics. Optimize