Speaker: Roberto Giuntini
Date and Time: Thursday, October 22nd 2020, 16:30-18:00, Amsterdam time.
Title: A quantum-like approach to Machine Learning
In this talk we propose a new quantum-like method for the binary classification applied to classical datasets. Inspired by the quantum Helstrom measurement, this approach allows one to define a new classifier, called Helstrom Quantum Classifier (HQC). This binary classifier (inspired by the concept of distinguishability between quantum states) acts on density matrices—called density patterns—that are the quantum encoding of classical patterns of a dataset. Different forms of quantum encodings will be presented and compared. We will then contrast the performance of HQC with that of a highly representative set of classical classifiers with respect to different classes of datasets. The experimental results show that HQC is very flexible and outperforms the other classifiers with respect to many (artificial and real-world) datasets. Finally, we will show that the performance of our classifier is positively correlated to the increase in the number of “quantum copies” of a pattern and the resulting tensor product thereof.