Abstract: Data Science is a field with immense breadth and depth. As our toolkit grows and we learn how to chain models and pipelines together, there's no limit to the time and complexity we can devote to over-engineering a solution. Many data scientists seem to be much better at learning algorithms and software libraries than they are at identifying good business cases to solve, and designing workflows that will enable them to work productively from start to finish.
In this workshop we look into the planning and design process of data science projects and explore strategies for building efficient workflows that lead to fast prototyping and seamless iterations of machine learning models. We also consider the human side of data science research and look at how increased understanding of our own cognitive processes can limit the impact that our biases and assumptions may have on our current and future work.
Bio: Cliff is a Senior Data Scientist and Instructor at Metis, where he teaches a 12-week immersive data science bootcamp covering machine learning, math and statistics, and the Python data science ecosystem. Previously he has worked on user segmentation analysis for Microsoft Office, trained natural language processing models for Cortana, built email classification and event extraction algorithms for Outlook, and automated demand forecasting for Amazon Fresh. He has also spent four years as a hedge fund quant in Chicago, after obtaining Master’s degrees in Statistics and Economics from the University of Chicago.