Loading…
Back To Schedule
Sunday, October 22 • 1:55pm - 2:40pm
Learning to learn Model Behavior: How to use "human-in-the-loop" to explain decisions.

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
The adoption of Machine Learning or Statistical Models in solving real-world problems has increased exponentially, but users still struggle to derive full potential of the predictive models. There is still a dichotomy between explainability and model performance while choosing the algorithm. Linear Models / Simple Decision Trees are often preferred over more complex models such as Ensembles or Deep Learning models when operationalizing models for ease of interpretation which often results in loss of accuracy.

But, is it necessary to accept a trade-off between model complexity and interpretability?Being able to interpret and explain a model globally faithfully helps in understanding feature contribution on predictions and model variability in a non-stationary environment. This enables trust in the algorithm which drives better collaboration and communication among peers. The need to understand the variability in the predictive power of a model in a human-interpretable way becomes even more important for complex models, e.g., text, image, machine translations.

In this talk, we demonstrate the usefulness of our Model Interpretation library (Skater) for evaluating models using interactiveness and usefulness of Jupiter environment and how it could help practitioners - analysts, data scientists, statisticians - understand the model behavior better without compromising on the choice of algorithm.

Speakers
avatar for Pramit Choudhary

Pramit Choudhary

Lead Data Scientist, DataScience.com
Pramit Choudhary is a Lead Data Scientist at Data Science.com. His focus is on effective ways of optimizing and applying classical (Machine Learning) and Bayesian design strategy to solve real-world problems. Currently, he is leading initiatives on figuring out better ways to explain... Read More →


Sunday October 22, 2017 1:55pm - 2:40pm PDT
Room D Room D