Ph.D. Proposal Oral Exam - Bradley Whitaker

Event Details
  • Date/Time:
    • Wednesday January 4, 2017
      1:30 pm - 3:30 pm
  • Location: Room 5244, Centergy
  • Phone:
  • URL:
  • Email:
  • Fee(s):
    N/A
Contact
No contact information submitted.
Summaries

Summary Sentence: Using Modified Sparse Coding as an Unsupervised Feature Extractor for the Binary Classification of Imbalanced Datasets

Full Summary: No summary paragraph submitted.

Title:  Using Modified Sparse Coding as an Unsupervised Feature Extractor for the Binary Classification of Imbalanced Datasets

Committee: 

Dr. Anderson, Advisor   

Dr. Rozell, Chair

Dr. Romberg

Abstract:

The objective of the proposed research is to explore the use of sparse coding as a tool for unsupervised feature learning to help in the classification of imbalanced datasets. Traditional sparse coding dictionaries are learned by minimizing the average approximation error between a vector and its sparse decomposition. As such, these dictionaries may overlook important features that are only present in the minority class. Without these features, it may be difficult to accurately classify between the two classes. To overcome this problem, this proposed work will explore novel modifications to the dictionary learning framework that will encourage dictionaries to include anomalous features. Sparse coding also inherently assumes that a vector can be represented as a sparse linear combination of a feature set. This proposal will address the ability of sparse coding to learn a representative dictionary when the underlying data has a nonlinear sparse structure. Finally, this work will apply the intuition gained from exploring dictionary learning in this manner to the binary classification of imbalanced datasets.

Additional Information

In Campus Calendar
No
Groups

ECE Ph.D. Proposal Oral Exams

Invited Audience
Public
Categories
Other/Miscellaneous
Keywords
Phd proposal, graduate students
Status
  • Created By: Daniela Staiculescu
  • Workflow Status: Published
  • Created On: Dec 23, 2016 - 10:13am
  • Last Updated: Dec 23, 2016 - 10:13am