The Secret Lives of Predictive Models
In the context of AI in general and learning predictive models in particular, predictability is usually considered a blessing. After all – that is the goal: build the model that has the highest predictive performance. The rise of ‘big data’ has in fact vastly improved our ability to predict human behavior thanks to the introduction of much more informative features. However, in practice things are more differentiated than that. In this talk Dr. Claudia Perlich will share some practical lessons when models had a surprising secret life and did something very different from what she thought she had asked them to do. As the creators of machine learning solutions it is our responsibility to pay attention to the often subtle symptoms and to let our human intuition be the gate keeper deciding when our models are ready to be released 'into the wild'.
Claudia Perlich joined Two Sigma as a Senior Data Scientist from Dstillery, where she served as a Chief Scientist (2010 to 2017). As a research staff member in the Data Analytics Research Group at the IBM Watson Research Center (2004 to 2010), she led teams that completed successfully in KDD Data Mining Competitions, designed and executed wallet/opportunity estimation models for IBM Sales using quantile regression, and worked on blog and Twitter analysis tools for marketing. Since 2011, Claudia has also worked as an adjunct professor teaching Data Mining in the M.B.A. program at the New York University Stern School of Business where she received her Ph.D. in Information Systems in 2004.
In the context of AI in general and learning predictive models in particular, predictability is usually considered a blessing. After all – that is the goal: build the model that has the highest predictive performance. The rise of ‘big data’ has in fact vastly improved our ability to predict human behavior thanks to the introduction of much more informative features. However, in practice things are more differentiated than that. In this talk Dr. Claudia Perlich will share some practical lessons when models had a surprising secret life and did something very different from what she thought she had asked them to do. As the creators of machine learning solutions it is our responsibility to pay attention to the often subtle symptoms and to let our human intuition be the gate keeper deciding when our models are ready to be released ‘into the wild’.
Claudia Perlich joined Two Sigma as a Senior Data Scientist from Dstillery, where she served as a Chief Scientist (2010 to 2017). As a research staff member in the Data Analytics Research Group at the IBM Watson Research Center (2004 to 2010), she led teams that completed successfully in KDD Data Mining Competitions, designed and executed wallet/opportunity estimation models for IBM Sales using quantile regression, and worked on blog and Twitter analysis tools for marketing. Since 2011, Claudia has also worked as an adjunct professor teaching Data Mining in the M.B.A. program at the New York University Stern School of Business where she received her Ph.D. in Information Systems in 2004.
