Model Validation and Interpretation

Join us to talk about how to streamline the process of initially defining and later re-defining your data model. What should engineers be doing? What should analysts and data scientists being doing? How can we all just get along?

### Speakers ###

Olivier Blais, Moov AI (https://moov.ai/en/)

Julien Beaulieu, Consultant

Interested in giving a talk? Reach out to us at matthieu.dimercurio@gmail.com

### Schedule ###

6:00 | Doors open

6:30 | First Presentation

Model Validation with Adversarial Examples, Olivier Blais

Advanced machine learning and AI models get more and more powerful. They also tend to become more complicated to validate and monitor.

To ensure that models work as promised, different techniques exist and are being created. One of which is to measure its sensitivity on worst case scenarios against adversarial examples. Olivier will present why this technique is awesome to validate deep learning models and how to use it.

7:15 | Second Presentation

Interpreting Complex Models, Julien Beaulieu

Julien will share different ways of making machine learning predictions more interpretable with a focus on tree-based approaches. He will demonstrate methods used to explain global model behavior across all data instances as well as methods for individual predictions. He will show how these methods work under the hood, what their strengths and weaknesses are, and how their outputs can be interpreted.

### Speaker Bios ###
Olivier Blais
Olivier is cofounder and Head of Data Science of Moov AI, a data science consulting company. He is co-holder of a patent for having created an advanced algorithm for assessing borrowing capacity. Olivier is a data science expert with extensive experience in supporting and implementing projects in companies in different sectors in their digital transformation: financial services, technology, aerospace, telecommunications and consumer products. He led the data team and implemented a data culture at Pratt & Whitney Canada, L’Oréal and GSoft.

Julien Beaulieu
Julien is a former digital marketer that recently transitioned into data science. He is currently pursuing his own personalized curriculum in order to learn state of the art techniques with the goal of helping companies make better decisions using data.

------------------------
Hosted by Dialogue