Machine Learning Bias

amrrs


Description:

We have been constantly told this statement “Computers don’t lie”. Yes in fact Computers don’t lie, but neither does it speak the truth. A computer does what its Master programs it to do. Similarly, A model wouldn’t lie unless the Machine Learning Engineer doesn’t want it to lie. ML Bias has become a global topic with AI getting into social life of a lot of us but the awareness among the tech community is still nascent that not all of us get it. This is an attempt to increase the awareness of what's ML Bias and how it's impacting!

Outline:

  • Recognizing the problem (with samples of Machine Learning Bias) - 5 mins
  • What's Machine Learning Bias (attempting to formulating a definition) - 5 mins
  • Definition of Fairness (understanding fairness and potential causes of bias) - 10 mins
  • Interpretable Machine Learning - 5 mins
  • Case Study (If time permits) - 5 mins

Meta

More Information: We evangelise Machine Learning and Data Science so much that almost every college fresher wants to get into this domain but we hardly take time to discuss what kind of adverse social impact these things bring with the current state of work. ML Bias/Ethics has been a topic in the US and Europe but we hardly pay attention to this (the same way we don't pay attention to Privacy) so I thought as a Data Science/ML Practitioner it's my duty to spread this awareness. and this submissions is one of the attempts for the same.

Motivation for the topic:

Quite some time back, I took part in a Kaggle Challenge analysis Kaggle Survey data and since then this topic has become one of my interest areas.

Motivation to submit to Pycon:

I recently gave this talk in Bangpypers (Bangalore Python user group) and the response was really good. so I thought it'd be a good idea to take it further to a larger base.

Prerequisites:

  • Basic understanding of Data Science and its applications
  • Knowledge of Data Science Workflow

Speaker Info:

Abdul Majed is an Analytics Consultant helping Organizations make sense some out of the massive - often not knowing what to do - data. Always amazed by Open Source and its contributors and trying to be one of them.

Organizer @ Bengaluru R user Group (BRUG) Organizer

Contributed to Open source by publishing packages on CRAN and PyPi

Writer @ Towards Data Science and DataScience+

Speaker Links:

https://datascienceplus.com/author/abdulmajed-raja/

Id: 1278
Section: Data Science, Machine Learning and AI
Type: Talks
Target Audience: Intermediate
Last Updated: