Seminar: Feature extraction by machine learning and its related physics Event
- Time:
- 12:00 - 13:00
- Date:
- 17 January 2020
- Venue:
- University of Southampton, Highfield Campus, Building 46, Room 5081
Event details
Shotaro Shiba-Funai, Okinawa Institute of Science and Technology will be delivering the first in a series of Turing at Southampton Seminars. These events are open to all.
Abstract:
Unsupervised machine learning is generally useful to extract features of input data. Since it can be regarded as a kind of information compression, some researchers suggest its similarity to coarse-graining and renormalization.
In this talk, we use spin configurations of Ising model as the input data and restricted Boltzmann machine (RBM) as the method of unsupervised learning. Then we look at what kind of features the machine extracts, using our method of "RBM flow". As a result, we can find an interesting similarity to renormalization and some coincidence with thermodynamics. However, we also discover apparent differences from renormalization and then argue why such phenomena occur by studying the parameter dependence in machine learning.