Publications

Towards Bipartisan Regulation of High Frequency Trading

  • Paper

  • Discusses the history and current state of high frequency trading in America, with particular emphasis on the positive and negative social and economic contributions of the industry as a whole. Proposes bipartisan responses which attempt to curtail abusive high frequency trading techniques while leaving the industry's considerable contributions to market liquidity relatively unharmed.

Lightweight Deconvolutional Neural Networks for Efficient Cloud Identification in Satellite Images

  • Paper

  • Poster

  • Develops a deconvolutional neural network architecture (drawing inspiration from a recent paper by Noh et al.) used to identify clouds in satellite imagery.

  • Originally authored as a capstone project for Stanford's course in Computer Vision and Deep Learning (CS 231N)

Towards Automatic Identification of Fake News: Headline-Article Stance Detection with LSTM Attention Models

  • Paper

  • Poster

  • Discusses the use of various neural network architectures built on top of Long-Short-Term-Memory models (LSTMs) for stance detection in the context of fake news. Taking an article as “ground truth”, we attempt to classify whether a given headline discusses, agrees, disagrees, or is unrelated to the article.

  • Submitted for consideration to Fake News Challenge.

  • Originally co-authored with Sahil Chopra and Saachi Jain as a capstone probject for Stanford's course in Natural Language Processing and Deep Learning (CS 224N)

Meet Percy: A CS 221 Teaching Assistant Chatbot

  • Paper

  • Poster

  • Discusses the implementation of an automated Piazza Teaching Assistant, designed to answer repetitive questions on Piazza online forums, allowing human teaching assistants to allocate more time to answering conceptual questions

  • Originally co-authored with Sahil Chopra and Rachel Gianforte as a capstone project for Stanford's Artificial Intelligence Course (CS 221)

Predicting Media Bias in Online News

  • Paper

  • Poster

  • Discusses the application of supervised and unsupervised methods to the quantification of media bias in various forms (selection bias, linguistic bias, etc…)

  • Originally co-authored with Noa Glaser as a capstone project for Stanford's Machine Learning Course (CS 229)

Bitcoin as Currency and Catalyst