عنوان

Algorithms for Discovering Collections of High-Quality and Diverse Solutions, with Applications to Bayesian Non-Negative Matrix Factorization and Reinforcement Learning

Number

TL54377

.Language of Text, Soundtrack etc

انگلیسی

Title Proper

Algorithms for Discovering Collections of High-Quality and Diverse Solutions, with Applications to Bayesian Non-Negative Matrix Factorization and Reinforcement Learning

General Material Designation

[Thesis]

First Statement of Responsibility

Masood, Muhammad Arjumand

Subsequent Statement of Responsibility

Doshi-Velez, Finale

Name of Publisher, Distributor, etc.

Harvard University

Date of Publication, Distribution, etc.

2019

Text of Note

201 p.

Dissertation or thesis details and type of degree

Ph.D.

Body granting the degree

Harvard University

Text preceding or following the note

2019

Text of Note

Machine Learning problems often admit a solution space that is not unique. When multiple feasible solutions exist, picking from a diverse, representative set may lead to better generalization and task-specific performance. While the emphasis of much of the literature has been on directly finding the `best' solution, we show that often a diverse set of near optimal solutions can be found which may be useful to practitioners and experts using machine learning models in decision making. This thesis investigates methods for obtaining a useful collection of solutions in specific models. Non-negative Matrix Factorization (NMF) is a popular data exploration tool and its Bayesian formulation is a promising approach for understanding uncertainty within this structure. We demonstrate that current approaches are lacking in the proper characterization of uncertainties and present novel techniques to provide model flexibility and improve the quality and speed of the inference. These techniques are applied to standard benchmark datasets for NMF as well as a curated medical dataset for understanding comorbidities in the Autism Spectrum Disorder (ASD). We show how a distinct collection of NMFs of nearly equal quality give rise to variability in interpretation of features and subsequent predictions. Finally, we present extensions of our diverse collection-based approach to the on-policy and off-policy Reinforcement Learning setting. Here, a completely new set of technical tools is required. In both on-policy and off-policy variants, we use diversity as a regularization feature in order to obtain a set of high-quality diverse policies. In addition to finding diverse policies in simulate-able multi-goal domains, we find a diverse set of policies designed to aid clinical decision making using ICU data for sepsis and hypotension management.

Subject Term

Artificial intelligence

Masood, Muhammad Arjumand

Doshi-Velez, Finale

Harvard University

Electronic name

p

[Thesis]

276903

a

Y