I am a PhD student in computational topology at the Topological Machine Learning Lab, UB (TML@UB) and at the HuPBA Lab funded by an FPU predoctoral research contract. Specifically I specialize in topological persistence and topological machine learning applied to the interpretability of neural networks. My research interests are (pure/computational) topology and geometry, category theory, algorithms, and artificial intelligence theory. Finally, I am supervised by Carles Casacuberta and Sergio Escalera.

During my undergraduate courses, I collaborated several times with the HuPBA group. There, I started my research career (which is, currently, in an early stage) with the paper Deep Multimodal Pain Recognition: A Database and Comparision of Spatio-Temporal Visual Modalities, where I learnt (and applied) several deep learning techniques to analyze our new dataset on state-of-the-art (machine learning and) computer vision neural networks. However, I noticed some time after the publication that I was more interested in mathematical content, so I started specializing in mathematics, especially in topological data analysis, which I applied to the study of neural network structures to predict neural network capacity to generalize to unseen data in my Bachelor's final thesis and to the improvement of neural networks' generalization capacities in my Master's thesis.

You can reach me on LinkedIn or directly in my email: rballeba@gmail.com.