Podcasts

---

Forget ROC Scores, What Metrics Do Your Stakeholders Care About?

Co-speaker: Christina Cai, Co-Founder & COO, Knowtions Research

How do you get buy-in from leadership to sponsor your ML project? How do you convince your stakeholders to put your ML models into production? And finally, how do you communicate the ROI of the project once it’s been deployed? While these are questions universal to any industry, they are particularly challenging to answer in the insurance industry because of its highly regulated and risk-averse nature. As such, more creative thinking is needed to convince stakeholders that your ML solutions can be trusted and bring value.

Video: https://www.youtube.com/watch?v=dpkf8avMBCQ

Taking a peek under the hood: Interpreting black box models

A model’s interpretability is just as important as its performance. In some industries, even more so. Unfortunately, some high performing models, like neural networks and ensemble methods, act more like black boxes. As practitioners, we are often asked to make a trade-off: interpretability or performance. Fortunately, as these complex models increase in popularity, there are ways to take a peek under the hood and interpret them. In this talk, we’ll present SHAP (Shapley Additive exPlanations) proposed by Lundberg et al., a model agnostic method to interpret machine learning models and how you can use it to explain your own models.

Video: https://www.youtube.com/watch?v=XyF21GMBdbk

The Whole is Greater Than the Sum of its Parts: Using Combinatorial Optimization to Create Synergy

What is the best route to get home? How much clothing should I pack for my vacation? Where do I seat Uncle Joe at the wedding reception so he's least likely to get into an argument with the wedding guests? These are all questions that can be answered using combinatorial optimization (CO). The key is recognizing that they are optimization problems and formulating them as such. An example--albeit contrived--would be "how do I create my wedding seating chart to maximize the amount of fun experienced by all guests?"  

While there are many everyday problems that can be solved using CO, these class of problems are also found in business. In this talk, we'll introduce combinatorial optimization problems and how to use Google's OR Tools library to solve them. We'll go over two business use cases and show how the Knapsack problem and constraint optimization can be used to maximize user enjoyment and team synergy.

Modelling user journeys: What will your customer do next?

Regardless of your business, being able to anticipate your users’ next action is a valuable advantage, whether that be a purchase, a view, or even a cancellation. Typical modelling approaches to predict users’ actions have focused on one specific action, e.g., conversion or churn. Here, we take a more holistic approach and don’t limit ourselves to one action. We model a users’ journey, so that we can not only anticipate a user’s action but also the one after that.

The Entrepreneurship Society, in partnership with Northeastern University, Rogers, BMO, CEO Global Network, EDC & BDC, hosted a conference all about AI & How it Will Affect the World We Live In

Panel Session: Common mistakes organizations make when implementing AI