Skip to Content

W51: Quantum Machine Learning Neural Network Design

200IJ

Sponsoring Units: DQI,GDSChair: Changhao LiSession Tags:
  • Focus

Thu. March 7, 3:48 p.m. – 4:00 p.m. CST

200IJ

Quantum machine learning (QML) is a computational paradigm which aims to use the principles of quantum mechanics to enhance the efficiency and performance of ML models for various learning tasks. Prior theoretical work has proven an unconditional expressivity separation, stemming from the quantum computer's use of quantum contextuality, between certain types of quantum and classical models for sequence learning tasks. This talk presents recent research exploring the role that quantum contextuality plays within QML models learning the latent distribution underlying genomic sequences. We discuss empirical results obtained using both synthetically generated and real world data sets, which test the effect that contextuality has on a QML model's abilility to efficiently capture long-range correlations in the data. This work aims to shed light on the sources of practical quantum advantages and better inform the search for application domains where such advantages may be realized.

Presented By

  • Teague Tomesh (Infleqtion)

Authors

  • Teague Tomesh (Infleqtion)
  • Pranav Gokhale (Infleqtion)
  • David Owusu-Antwi (Infleqtion)