I'm a CS Ph.D. candidate in Georgia Tech's School of Interactive
Computing and I work in Prof. Byron Boots' Robot Learning Lab.
I graduated from Harvey Mudd College in May 2016 with a B.S. in Physics and concentrations in Computer Science and Political Studies.
My general research focus is on exploring connections between machine learning and quantum information theory. I am especially interested in quantum graphical models, which are similar to classical probabilistic graphical models, but use density matrices (instead of probability vectors) as the fundamental state representation.
I also enjoy learning and thinking about philosophy, mathematics, physics, and thought-experiments/paradoxes in general (my favourites include the Sleeping Beauty problem, the Repugnant Conclusion, and of course, crazy variations of the trolley problem -- but this list changes regularly). In my spare time, I enjoy scuba diving, swing/salsa dancing, listening to podcasts, hiking, and traveling to new places.
I have recently become familiar with the effective altruism movement, and I find their commitment towards improving well-being and reducing suffering through evidence-based methods very compelling. You can learn more about the movement here. If you are convinced by the goals of the movement, please consider donating to an effective charity here.
Quantum Graphical Models (QGMs) generalize classical graphical models by adopting the formalism for reasoning about uncertainty from quantum mechanics. Unlike classical graphical models, QGMs represent uncertainty with density matrices in complex Hilbert spaces. Hilbert space embeddings (HSEs) also generalize Bayesian inference in Hilbert spaces. We investigate the link between QGMs and HSEs and show that the sum rule and Bayes rule for QGMs are equivalent to the kernel sum rule in HSEs and a special case of Nadaraya-Watson kernel regression, respectively. We show that these operations can be kernelized, and use these insights to propose a Hilbert Space Embedding of Hidden Quantum Markov Models (HSE-HQMM) to model dynamics. We present experimental results showing that HSE-HQMMs are competitive with state-of-the-art models like LSTMs and PSRNNs on several datasets, while also providing a nonparametric method for maintaining a probability distribution over continuous-valued features.Hilbert Space Embeddings, Hidden Quantum Markov Models, Kernels
As social robots increasingly enter people’s lives, coherence of personality is an important challenge for longterm human-robot interactions. We extend an architecture that acquires dialog through crowdsourcing to author both verbal and non-verbal indicators of personality. We demonstrate the efficacy of the approach through a four-day study in which teams of participants interacted with a social robot expressing one of two personalities as the host of a competitive game. Results indicate that the system is able to elicit personality-driven language behaviors from the crowd in an incremental and ongoing way and produce a coherent expression of that personality during face-to-face interactions over time.Human-Robot Interaction
In the Story Cloze Test, a system is presented with a 4-sentence prompt to a story, and must determine which one of two potential endings is the 'right' ending to the story. Previous work has shown that ignoring the training set and training a model on the validation set can achieve high accuracy on this task due to stylistic differences between the story endings in the training set and validation and test sets. Following this approach, we present a simpler fully-neural approach to the Story Cloze Test using skip-thought embeddings of the stories in a feed-forward network that achieves close to state-of-the-art performance on this task without any feature engineering. We also find that considering just the last sentence of the prompt instead of the whole prompt yields higher accuracy with our approach.Story Cloze Test, Natural Language Processing
Hidden Quantum Markov Models (HQMMs) can be thought of as quantum probabilistic graphical models that can model sequential data. We extend previous work on HQMMs with three contributions: (1) we show how classical hidden Markov models (HMMs) can be simulated on a quantum circuit, (2) we reformulate HQMMs by relaxing the constraints for modeling HMMs on quantum circuits, and (3) we present a learning algorithm to estimate the parameters of an HQMM from data. While our algorithm requires further optimization to handle larger datasets, we are able to evaluate our algorithm using several synthetic datasets. We show that on HQMM generated data, our algorithm learns HQMMs with the same number of hidden states and predictive accuracy as the true HQMMs, while HMMs learned with the Baum-Welch algorithm require more states to match the predictive accuracy.Hidden Markov Models, Machine Learning, Quantum Physics
This paper introduces the use of compressed sensing for autonomous robots performing environmental mapping in order to reduce data collection, storage, and transmission requirements. A prototype robot sends data collected over adaptively updated straight-line paths to a server, which reconstructs an image of the environment variable using SplitBregman iteration. The amount of data collected is only 10% of the amount of data in the final map, yet the relative error is only 20%.Compressed Sensing, Autonomous Control, Robots
Shoot me an email if you have any questions or if you just want to chat!