Daniel González Cortés
Data Scientist & AI Researcher
Paris, France
Research Papers
Disclosure dynamism: TCFD aligned climate claims of UK corporates
This publication applies the ClimateBERT machine learning model to analyze corporate climate claims from FTSE 100 companies over eight years. The findings highlight a significant increase in climate-related disclosures, demonstrating the impact of NLP-driven analysis on corporate transparency and policy evaluation.
Portfolio construction using Explainable Reinforcement Learning
An explainable reinforcement learning model is introduced for portfolio management, enhancing transparency in financial machine learning. Applied to a custom CAC-40 trading environment, the model dynamically adapts to market changes and outperforms an equally weighted portfolio in out-of-sample tests.
Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to Financial Time Series
This study presents a clustering framework that utilizes autoencoders to enhance the representation of financial time series data, leading to improved predictive model performance. When applied to major financial indices, this approach enhances clustering accuracy, providing deeper insights that support more effective investment strategies and risk management.