Loading…
Back To Schedule
Friday, September 24 • 10:00am - 12:00pm
Parallel Workshop 1 : Understand and build trust in your machine learning models by Google (Limited Seats) - 120 Minutes

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

About the Workshop:
Topic: Understand and build trust in your machine learning models


As investments in AI/ML grow, it has become important to build end-user trust in machine learning
models. AI technology suffers from what is called a “black box” problem. Stakeholders are often reluctant
to trust machine learning projects because they don’t understand what they do. It’s hard for
decision-makers to relinquish control to a seemingly mysterious machine learning model, especially when
it’s responsible for crucial decisions. This is more pronounced in the case of deep learning and artificial
neural network approaches, which contain many hidden layers that learn through pattern recognition.
To ensure the success of AI/ML programs and drive adoption, it has therefore become important to
enhance transparency with human-interpretable explanations of machine learning models, and to have
the ability to ensure that models are kept current, with predictions that are in line with ground truths.
In this workshop, we shall look at some of these concepts, associated tools and frameworks, and
understand them with the help of some demonstrations. You shall be able to follow along for some of the
demonstrations with Google Developers Codelabs, or try them later.


Agenda:
Time & Topic
09:30 am - 09:45 am The need for trust and transparency in ML models
09:45 am - 10:15 am Explanations for models on structured and image data; Understanding performance using the What-if Tool
10:15 am - 10:45 am Automation in ML and explainability in AutoML Brief overview of Neural Architecture Search
10:45 am - 11:00 am Keep models current with ML pipelines Example with the KubeFlow Pipelines SDK
11:00 am - 11:15 pm Increased transparency for ML models with model cards
11:15 pm - 11:30 pm Conclusion and Q&A

Speakers
avatar for Prasanna Keny

Prasanna Keny

Customer Engineer, Analytics, Google india
Prasanna is a Customer Engineer at Google Cloud. As a data, analytics and machine learning specialist, he helps enterprises drive real business outcomes and realise value from their analytics initiatives. His experience in roles spanning advisory services, consultative sales, program... Read More →


Friday September 24, 2021 10:00am - 12:00pm IST