Kang-Hao Peng - MS Thesis Defense

Thursday, August 4, 2016
1:30 p.m.
2328 AVW
Melanie Prange
301 405 3686
mprange@umd.edu

Announcement: MS Thesis Defense

Name: Kang-Hao Peng
 
Committee: Professor Rama Chellappa (chair), Professor Joseph JaJa, Professor Prakash Narayan, Professor Richard La
 
Date/Time: August 4, 2016, 1:30 pm
 
Location: 2328 AVW
 
Title: Channel Capacity-Based RBM Neural Networks
 
Abstract: 
(Deep) neural networks are increasingly being used for various computer vision and pattern recognition tasks due to their strong ability to learn highly discriminative features.  However, quantitative analysis of their classification ability and design philosophies is still nebulous. In this work, we use information theory to analyze channel capacity of concatenated restricted Boltzmann machines (RBMs) and propose a channel capacity-based RBM neural networks (CC-RBM). We develop a novel pre-training algorithm to maximize the mutual information between RBMs. Extensive experimental results on various classification tasks show the effectiveness of the proposed approach.

Audience: Graduate  Faculty 

remind we with google calendar

 

March 2024

SU MO TU WE TH FR SA
25 26 27 28 29 1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31 1 2 3 4 5 6
Submit an Event