Title:
Feedback Coding for Efficient Interactive Machine Learning

dc.contributor.advisor Rozell, Christopher J.
dc.contributor.author Canal, Gregory Humberto
dc.contributor.committeeMember Davenport, Mark
dc.contributor.committeeMember Bloch, Matthieu
dc.contributor.committeeMember Nowak, Robert
dc.contributor.committeeMember Xie, Yao
dc.contributor.department Electrical and Computer Engineering
dc.date.accessioned 2021-09-14T17:18:51Z
dc.date.available 2021-09-14T17:18:51Z
dc.date.created 2021-08
dc.date.issued 2021-07-14
dc.date.submitted August 2021
dc.date.updated 2021-09-14T17:18:52Z
dc.description.abstract When training machine learning systems, the most basic scenario consists of the learning algorithm operating on a fixed batch of data, provided in its entirety before training. However, there are a large number of applications where there lies a choice in which data points are selected for labeling, and where this choice can be made “on the fly” after each selected data point is labeled. In such interactive machine learning (IML) systems, it is possible to train a model with far fewer labels than would be required with random sampling. In this thesis, we identify and model query structures in IML to develop direct information maximization solutions as well as approximations that allow for computationally efficient query selection. To do so, we frame IML as a feedback communications problem and directly apply principles and tools from coding theory to design and analyze new interaction selection algorithms. First, we directly apply a recently developed feedback coding scheme to sequential human-computer interaction systems. We then identify simplifying query structures to develop approximate methods for efficient, informative query selection in interactive ordinal embedding construction and preference learning systems. Finally, we combine the direct application of feedback coding with approximate information maximization to design and analyze a general active learning algorithm, which we study in detail for logistic regression.
dc.description.degree Ph.D.
dc.format.mimetype application/pdf
dc.identifier.uri http://hdl.handle.net/1853/64945
dc.language.iso en_US
dc.publisher Georgia Institute of Technology
dc.subject Feedback coding
dc.subject Machine learning
dc.subject Information theory
dc.subject Interactive learning
dc.subject Active learning
dc.title Feedback Coding for Efficient Interactive Machine Learning
dc.type Text
dc.type.genre Dissertation
dspace.entity.type Publication
local.contributor.advisor Rozell, Christopher J.
local.contributor.corporatename School of Electrical and Computer Engineering
local.contributor.corporatename College of Engineering
relation.isAdvisorOfPublication 301419ce-552b-4510-b374-2eecfed3f42a
relation.isOrgUnitOfPublication 5b7adef2-447c-4270-b9fc-846bd76f80f2
relation.isOrgUnitOfPublication 7c022d60-21d5-497c-b552-95e489a06569
thesis.degree.level Doctoral
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
CANAL-DISSERTATION-2021.pdf
Size:
31.33 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
LICENSE.txt
Size:
3.87 KB
Format:
Plain Text
Description: