Interview with Jeremiah Johnson, Professor, University Of New Hampshire - Speaker at Global Data Science Conference - Oct - 2017 Posted on : Sep 25 - 2017

We feature speakers at Global Data Science Conference - Oct - 16 - 18 2017 - Boston - MA to catch up and find out what he or she is working on now and what's coming next. This week we're talking to Jeremiah Johnson, Professor, University Of New Hampshire (Topic : Cheating At Deep Learning: Quick Prototyping Using Keras)

Interview with Jeremiah Johnson

1. Tell us about yourself and your background.

I am Professor of Data Science at the University of New Hampshire, where I do research in machine learning and mathematics. I direct the university's Bachelor of Science in Analytics program. 

2.  What have you been working on recently?

I work on applying machine learning and neural networks in a variety of settings, from medical image analysis to detection of artistic style in paintings. I also work to expand our understanding of the mathematical and theoretical nature of neural networks, so that we can keep up the incredible progress made in this area in the past few years.

3. Tell me about the right tool you used recently to solve a problem?

I use a variety of tools, it really just depends on the job at hand. My favorites are the usual Python ML stack (NumPy, SciPy, Matplotlib, Scikit-Learn), Keras, and Tensorflow. 

4. Where are we now today in terms of the state of Data Science, and where do you think we’ll go over the next five years?

Data Science is making inroads in every industry. I expect this trend will continue and even accelerate, as companies acquire more data and technologies that make use of machine learning and artificial intelligence proliferate.

5. What are some of the best takeaways that the attendees can have from your "Cheating At Deep Learning: Quick Prototyping Using Keras " talk?

The faster you can try out new models, the quicker you will find good results. Keras has a simple, intuitive API that makes it so easy to quickly build models (convolutional neural networks, recurrent neural networks, and more) that you’ll feel like you’re cheating.