DataSciPy: State of the Art of Deep Neural Networks - Handwriting Recognition

Deep learning is being applied to every field that data scientists can get their hands on, but it still has its limitations. The field of automated handwriting recognition is pushing the boundaries of deep learning because documents are being digitized at a faster rate than humans can keep up with.

In this presentation, Danny Hanson will discuss his capstone research on handwriting recognition and its integration into the crowdsourcing platform Zooniverse. He will discuss the state of the art in deep neural networks and how to collect appropriate data.

Danny will talk about where the solutions for MNIST fail to generalize and how handwriting and speech are quite similar from a computer's perspective. Danny will include an interactive session using Jupyter notebook to show some aspects of his model.