Resume/Curriculum Vitae
David Weiss
djweiss AT princeton DOT edu
www.djweiss.info
3-N-14 Green Hall
Princeton University
Princeton, NJ
Education
May 2007
Princeton University
A.B. in Computer Science (Highest Honors), Certificate in
Neuroscience
GPA: 3.74 (Overall), 3.85 (Major)
Awards
- Senior Thesis Prize (2007), Department of Computer Science, Princeton University
- Sigma Xi Book Award (2007), Princeton Chapter, Sigma Xi Association
- Sigma Xi Membership Nomination (2007), Princeton Chapter, Sigma Xi Association
- 2 Place Overall & 2 Place Engineering (2007), Princeton Undergraduate Research Symposium
- Honorable Mention (2007), Outstanding Undergraduate Award, Research Computing Association
- Gold Award (2003), Humor, Scholastic Art & Writing Awards
Professional Experience
July 2007 - Present
Research Specialist, Princeton Computational Memory
Lab
Department of Psychology, Princeton University
- Derived a variational Expectation Maximization algorithm to perform approximate inference in a matrix factorization model with non-conjugate priors (senior thesis)
- Developed a Matlab software suite for running distributed “searchlight” functional MRI brain mapping analyses, as well as resampling methods for nonparametric hypothesis testing
Summer 2006
Research Assistant, Princeton Computational Memory
Lab
Department of Psychology, Princeton University
- Developed for the Princeton Multi-Voxel Pattern Analysis (MVPA) Toolbox for Matlab, an open- source neuroimaging analysis suite
Summer 2005
Intern, Laboratory for Computational Motor
Control
Department of Biomedical Engineering and Neuroscience, Johns Hopkins
University
- Designed and implemented a software model of Macaque musculature capable of simulating muscle spindle activity
- Simulated sensory afferents to M1 direction-selective neurons and used a generalized linear model (GLM) to fit simulated activity to recorded neural spike trains during arm movements
April - June 2003
Intern, Haptics Exploration Laboratory
Department of Mechanical Engineering, Johns Hopkins University
- Designed software to simulate cutting forces in a 4D soft tissue simulation with tactile feedback
- Work was presented at the 12 Annual Medicine Meets Virtual Reality conference in 2004
Summers 2001–2004
Faculty, Computer Science
The Park School Science Camps, Brooklandville, MD
- Taught and designed two separate computer science curricula (for ages 8-10 and 11-13) to 8 classes of children over a 6 week period
Selected Research Projects
Fall 2006 – Fall 2007
Netflix Prize
Along with Lester Mackey and David Lin, founded the team “Dinosaur Planet” to compete in the Netflix
Prize competition, a contest in collaborative filtering with a $1 million Grand Prize and a $50k yearly
Progress Prize.
- Implemented a C framework for recommendation engines capable of efficiently handling the 100 million training examples in the Netflix dataset
- Implemented a model blending approach combining many different algorithms ranging from matrix factorizations (NMF, RMF, SVD, etc.), to clustering (KNN, GMM, etc.), to undirected graphical models (RBM, CRBM, etc.).
- Final score in 2007: 8.00% improvement over the Netflix Cinematch Algorithm (5 place overall), and a combined score of 8.38% after collaborating with Team Gravity (2 place overall)
Fall 2006 – Spring 2007
Probabilistic Additive Component Analysis
Senior thesis. Proposed, designed, and implemented a generative model of functional MRI datasets to
facilitate decoding of cognitive state from images of neural
activity.
- Applied maximum a posteriori estimation to fit the latent variables of the PACA model to a real world dataset, and evaluated the model using reconstruction error and classification accuracy
- Demonstrated that the PACA model outperforms two other popular matrix-factorization based dimensionality reduction methods, Principal Component Analysis (PCA) and Non-negative Matrix Factorization (NMF)
Spring 2006
“Playing Expert Rock, Paper, Scissors”
COS511
(Foundations of Machine Learning) Final Project. Adapted the online
learning algorithm in Hemold & Schapire (1997) to play games using a
game theoretic approach, with a specific application to the popular
game “Rock, Paper, Scissors.”
- Designed a website (link here) where the public can play against the algorithm online using a graphical interface
- Final score in 2006: 114 wins and 47 losses against humans
Fall 2005
“Reading Wavelets, Reading Minds”
Junior
independent work. Investigated the use of the Discrete Wavelet
Transform (DWT) to efficiently generate a large set of spatial
features for the classification of human functional MRI datasets.
Posters & Publications
- PDFQuamme, J.R., Weiss, D.J., & Norman, K.A. “Pattern classification of fMRI retrieval states in recognition memory.” Poster presented at the Annual Meeting of the Psychnonomic Society, Long Beach, CA (Nov, 2007).
- PDFWeiss, D.J., Blei, D.M., & Norman, K.A. “Probabilistic Additive Component Analysis: A Latent Variable Model for Dimensionality Reduction of Human fMRI Datasets.” Poster presented at the Princeton Undergraduate Research Symposium, Princeton, NJ (May, 2007).
- PDFWeiss, D.J. ''Probabilistic Additive Component Analysis: A Latent Variable Model for Dimensionality Reduction of Human fMRI Datasets.'' Senior Thesis, 2007.
- PDF Weiss, D.J., & Okamura, A.M. "Haptic Rendering of Tissue Cutting with Scissors in a Virtual Environment," Medicine Meets Virtual Reality 12, J.D. Westwood, et al. (Eds.), IOS Press, 2004, pp. 407-409.