Stephen Fitz

Self-introduction message

I'm an American scientist working in the field of Artificial Intelligence. My main research interests are Deep Neural Networks, Representation Learning, Algebraic Topology of Embedding Spaces, Computational Linguistics / NLP, and Reinforcement Learning. Additionally, I'm excited about the application of methodology and computational techniques from Neural Machine Learning and Mathematical Data Science to research in areas outside of mainstream AI, such as Physics, Biology, Neuroscience, as well as Social Sciences and Humanities. Intersection of Deep Learning and Applied Psychology is an area of interest I'm currently involved in as well. I'm also interested in Economics, Law, and Entrepreneurship. Outside the Academia, I have been involved in applications of AI in various industries. I was a founding member of the Chicago Innovation Exchange (center for new technology startups in Chicago, sponsored by the Booth School of Business at the University of Chicago) where I advised companies on the use of AI and data science in various domains, and I have been applying AI techniques to diverse problems including algorithmic trading. I have worked in technology sector positions including software development, consulting, industrial research, as well as executive appointments.
One of my main hobbies has been frequent travels around the world (I lived in several countries and visited more than twenty, spanning multiple continents). I particularly enjoy staying in Tokyo, and it has been my frequent travel destination.


Background and career
I studied for my undergraduate degree (BA) at Columbia University (New York, USA), and for my graduate degrees (MPhil, MA, PhD) at University of Cambridge (Cambridge, UK) and University of Chicago (Illinois, USA).

I taught various courses on AI, programming, data science, computational linguistics, statistics, and mathematics at universities in the United States and in Japan. I am currently teaching courses on Artificial Intelligence at Keio University, and I'm a Project Assistant Professor at the Keio Global Research Institute.

Research achievements

  • "Neural Networks And Deep Learning: A Paradigm Shift In Information Processing, Machine Learning, And Artificial Intelligence", Book Chapter (Handbook of Alternative Finance), Palgrave Macmillan
  • "The Use of Psychometrics and Artificial Intelligence in Alternative Finance", Book Chapter (Handbook of Alternative Finance), Palgrave Macmillan
  • "The Shape of Words: Topological Features of Natural Language Representations in Vector Spaces", Harvard Data Science Review.
  • "Evaluation of Representation Methods for Graph Convolutional Sequence-to-Sequence Models", Empirical Methods in Natural Language Processing (submission).
  • "Algebraic Invariants of Adversarial Examples in Deep Neural Vision Systems", International Conference on Machine Learning (submission).
  • "Autonomous Vehicle Navigation from Natural Language Instructions and Raw Sensory Data via a Deep Recurrent Encoder-Decoder Model", unpublished research project at Toyota Technological Institute.
  • "KOGITO - 2L Lexical Acquisition Using Distributed Vector Space Representations of Linguistic Data", applied industry research project (including a publication of a commercial software system).
  • "Deep Grammar Correction", unpublished applied research project at University of Cambridge resulting in the development of a GEC system.