Presented By O’Reilly and Intel AI
Put AI to work
Sep 4-5, 2018: Training
Sep 5-7, 2018: Tutorials & Conference
San Francisco, CA

Trustworthiness of machine learning applications

Mayukh Bhaowal (Salesforce)
4:00pm-4:40pm Friday, September 7, 2018
AI Business Summit, AI in the Enterprise
Location: Continental 7-9 Level: Beginner
Secondary topics:  Ethics, Privacy, and Security

Who is this presentation for?

  • Everyone will find value in this session.

Prerequisite knowledge

  • Familiarity with a learning software or machine learning

What you'll learn

  • Understand the importance of interpretability and the trade-offs among various approaches of making machine learning models interpretable
  • Explore the push-and-pull dynamics between interpretability and model accuracy

Description

Machine learning (ML) software differs from traditional software in the sense that outcomes are not based on a set of hand-coded rules and hence not easily predictable, since the behavior of such software changes over time based on data and feedback loops. Salesforce Einstein cares deeply about building trust and confidence in intelligent software programs. Why does a particular email have a higher likelihood of being opened than another? What are the shapes and patterns in the dataset that lead to certain predictions? And can such insights be actionable?

As machine learning pervades every software vertical and is increasingly used to automate decisions, model interpretability becomes an integral part of the ML pipeline and can no longer be an afterthought. In the real world, the demand for being able to explain a model is rapidly gaining on model accuracy and other model evaluation metrics. It all starts with cultivating the idea that trustworthiness goes hand in hand with interpretability and is something that needs to be carefully thought through from the get-go.

Mayukh Bhaowal discusses the steps Salesforce Einstein is taking to make machine learning more transparent and less of a black box. Mayukh explains how interpretability fits into the ML data pipeline, what Salesforce Einstein learned while trying different approaches, and how it has helped drive wider adoption of ML software.

Photo of Mayukh Bhaowal

Mayukh Bhaowal

Salesforce

Mayukh Bhaowal is a director of product management at Salesforce Einstein working on automated machine learning. Previously, Mayukh worked at startups in the domain of machine learning and analytics. He served as head of product of ML platform startup Scaled Inference, backed by Khosla Ventures, and led product at ecommerce startup Narvar, backed by Accel. He was also a principal product manager at Yahoo and Oracle. Mayukh holds a master’s degree in computer science from Stanford University.