Talks

A Brief Introduction to Hyperparameter Tuning

Toronto Womxn in Data Science Conference, Toronto, March 2nd, 2020

Hyperparameter optimization is a key step in the data science pipeline which aims to identify the hyperparameters that optimize model performance. This talk walks through the process of hyperparameter tuning using a case study of a sepsis prediction model. Covers the most popular techniques for hyperparameter optimization including grid search, random search, sequential mode-based optimization, and Bayesian optimization

Algorithmic Bias in Machine Learning

PyCon Canada, Toronto, November 16th, 2019

Machine learning algorithms are susceptible to both intentional and unintentional bias. Relying on biased algorithms to drive decisions can lead to unfair outcomes that have serious consequences affecting underrepresented groups of people. In this talk, we’ll walk through examples of algorithmic bias in machine learning algorithms, explore tools (in Python) that can measure this bias, and discuss good ethics and software engineering strategies to mitigate bias in machine learning algorithms.

How to Build a Clinical Diagnostic Model in Python

PyCon US, Cleveland Ohio, May 3rd 2019

Explores how machine learning can be used to build automatic clinical daignostic tools using data from the electronic medical record. A large part of the talk focuse on tools and techniques involved in pre-processing complex medical data, and strategies for evaluating model results.

A Brief Overview of Hyperparameter Optimization

PyData DC, Washington DC, November 18th, 2018

Discusses hyperparamter optimization in a clnical context. Investigates the most popular optimization techniques including: 1) grid search, 2) random search, 3) sequential model-based optimization, and 4) Bayesian optimization. See slides here.

Buildling a Recommender System from Scratch (tutorial)

PyData DC, Washington DC, November 16th, 2018

Walks through how to build a recommendation system using data from the MovieLens dataset. Covers various topics on recommenders such as collaborative and content-based filtering, implicit and explicit feedback, and the cold start problem.

How to Design and Build a Recommendation Pipeline in Python

PyCon Canada, Toronto Ontario, November 10th, 2018

Delves into the process of designing and building a recommendation system pipeline. Covers techniques including data cleaning and normalization, hyperparameter tuning, model training and fitting, as well as quantitative and qualitative model evaluation.

An Introduction to Recommendation Systems (tutorial)

PyLadies Meetup, Vancouver BC, August 30th, 2018

Walks through how to build a recommendation system using data from the MovieLens data. Covers various typics on recommenders including collaborative and content-based filtering, implicit and explicit feedback, and the cold start problem.