thanks for visiting
Henry Wolf . AI
Henry is a PhD Candidate using deep learning to study the cognitive science of reading and dyslexia at the University of Connecticut.
Henry works as a Senior Data Scientist and reports directly to the Chief Research Officer at the Advertising Research Foundation.
Henry enjoys international travel. He speaks English, Japanese, and Mandarin Chinese.
This project is an attempt to replicate a paper that used convolutional neural networks to replicate the performance of baboons in a lexical decision task. This was used as an example during a "Deep Learning 101" presentation. The talk introduced deep learning in TensorFlow to professors and graduate students in the Neurobiology of Language program at the University of Connecticut.
These models are use to determine the number of strokes in Japanese kanji. The primary model consists of a convolutional neural network, but a second model uses only a feedforward network.
This purpose of this project was to determine interest in learning languages by scraping Twitter and generating visualizations with regards to Chinese, English, and Japanese. This project could be improved by increasing the amount of data collected to include a larger time period, as overlapping topics are likely influenced by what is trending.
This is currently a two-part series on the basics of data cleaning and visualization in R. It was prepared and presented for graduate students at the University of Connecticut Department of Psychology. Tutorial One explains some of the functions available in dplyr for cleaning data. Tutorial Two goes through some of the plotting available in ggplot2. Bad jokes come free of charge.