In this talk, after a streamlined historical introduction to Artificial Intelligence, at first, I will summarize recent advances in our understanding about information processing by modern neural networks in the shallow limit: by using archetypical models for “pattern recognition” and “machine learning”, I will discuss two mathematical approaches that our group developed to tackle this problem, the former -Guerra’s interpolation- more probabilistic in its nature, the latter -PDE techniques- more analytical. Then I will close this talk by showing how the know-how resulting from such theoretical investigations translates into practical computational recipes, useful in applications, with a particular emphasis on problems related to cancerogenesis.<\/p>\n\n\n\n
Minimal Bibliography In this talk, after a streamlined historical introduction to Artificial Intelligence, at first, I will summarize recent advances in our…<\/p>\n
<\/strong>Agliari, E., Alemanno, F., Barra, A., & Fachechi, A. (2020). Generalized Guerra\u2019s interpolation schemes for dense associative neural networks. Neural Networks, 128, 254-267.
Agliari, E., Alemanno, F., Barra, A., Centonze, M., & Fachechi, A. (2020). Neural networks with a redundant representation. Physical review letters, 124(2), 028301.
Alemanno, F., et al. (2023) Quantifying heterogeneity to drug response in cancer\u2013stroma kinetics. Proceedings of the National Academy of Sciences 120.11: e2122352120.<\/p>\n","protected":false},"excerpt":{"rendered":"