Josh Silverberg

; About Me Experience Projects Contact

Projects

Tweet stock sentiment prediction

  • Creating market sentiment prediction program that scrapes recent tweets to classify any stock as bullish, bearish, or neutral with approximately 50% accuracy.
  • Implemented using custom and out-of-box prediction models from Tensorflow and scikit-learn.
  • Deploying system to AWS Lambda for easier access and faster computation.

    • Using Saliency Mapping to Visualize and Interpret Text Classifiers

  • Worked with interdisciplinary team of Masters and PhD students to create visualizations to interpret and understand text classification models.
  • Distilled complex NLP models into 1st and 2nd order linear approximations and compared these to transformer attention mappings.
  • Evaluated each approximation and visualization method and summarized results in symposium poster.

    • Styled text generation

  • Led NLP project team working on using RNNs and transformers to generate text in the styles of particular authors.
  • Coordinated and contributed to development of mutliple sucessful RNN and transformer-based models.

    • Conversational chatbot

  • Creating chatbot that uses seq2seq architecture to intelligently respond to user prompts.
  • Implemented using custom Tensorflow LSTM models.

    • Automatic styled reddit comment generation

  • Initiated, planned, and executed a project to compare common RNN-based models on their ability to generate Reddit comments in styles of three subreddits in response to a prompt.
  • Compiled and synthesized almost 100 academic papers into executive summaries in the meta-research phase.
  • Modified code from an ACL-2019 paper on LSTM-based seq2seq comment generation with user feature embedding using TensorFlow.

    • Double-ended sorts library

  • C++ library of two-sided versions of common sorts with templated support for unspecified containers and datatypes.