(Robert) Dustin Wehr

Recently graduated with a PhD in Computer Science from the University of Toronto, under the supervision of Stephen Cook and Alasdair Urquhart. Did some teaching. Now taking an indefinite leave from academia, but continuing to work on my thesispassion project, and other software stuff from home. As an MSc student I worked in Computational Complexity Theory.

To reach me: dustin.[my last name]@gmail.com.

Currently relying on linkedin for my resume...

Main project (2017)

Logic for Progress

Research stuff

My PhD thesis and related files.

Final May 2014 draft of Challenges and examples of rigorous deductive reasoning about socially-relevant issues, to be presented at Trends in Logic XIV. Also Thesis proposal from Feb 2014.

Paper proving the easier of the two main conjectures posed by Anna Gál, Michal Koucký, and Pierre McKenzie in their 2008 paper "Incremental branching programs":
Lower bound for deterministic semantic-incremental branching programs solving GEN
The harder problem, proving superpolynomial lower bounds for nondeterministic semantic-incremental branching programs, is still open.

Publications

The first publication listed below is a journal paper that contains all the results from the two conference papers that follow it in the list.

Fractional Pebbling and Thrifty Branching Programs, FSTTCS 2009, Mark Braverman, Stephen Cook, Pierre McKenzie, Rahul Santhanam, Dustin Wehr

The main results of my masters paper (arXiv) are the (tight) fractional pebbling lower bound (also in above FSTTCS and ACM publications) and the (tight) lower bound for deterministic thrifty branching programs solving the DAG evaluation problem (also in the ACM publication). It also contains some extensions of results from the above publications, including a generalization of the definition of thrifty branching program that greatly expands the number of input variables that a program may query.

Here is a writeup of an unpublished result that we mentioned in the ACM journal paper:
Exact size of smallest minimum-depth deterministic BPs solving the tree evaluation problem

Some school projects

Using Restricted Boltzmann Machines for recommendations (i.e. collaborative filtering): a description and some small improvements on the influential work of R Salakhutdinov, A Mnih, G Hinton (one of the key algorithms used by the winners of the Netflix Prize). This is a report on a course project with my friend Wesley George for Geoff Hinton's graduate course Introduction to Machine Learning.

A detailed statement and proof of Büchi's Theorem, which gives a relationship between Monadic Second Order Logic and finite automata on infinite words. This is a report I did during my undergrad for Prakash Panangaden's graduate course Formal Verification.

Hobby stuff

Firefox extension mingaling.