🎯 AI Fairnees #4: The Exclusion Bias
‼️ Today we're talking about a sneaky bias that can be introduced during data pre-processing: Exclusion Bias.
Exclusion bias happens when we remove certain features from our data, thinking they're irrelevant.
This can include things like null values, outliers, or other extraneous data points.
But if we're not careful, that removal process can lead to underrepresentation of certain features when the data is applied to a real-world problem, which can result in a loss of the true accuracy of our data.
Here's an example: let's say you're comparing referral rates from the English and Sinhala versions of a website. 98% of the clicks come from the English version, and only 2% come from the Sinhala version.
It might be tempting to leave out the 2% and focus on the English data, but if you do that, you might miss out on the fact that the Sinhala clicks have a higher conversion rate.
That's exclusion bias, and it can lead to some seriously misleading results.
So the next time you're working on a machine learning project, be careful not to exclude any features without a good reason.
It might seem like a small thing, but it can have a big impact on your results.
Hopefully you liked this video 💚
🔥 Subscribe to Bitswired
👍🏽 Leave us a like/comment to support us
💬HASHTAGS:
#bias #fairness #machinelearning #ai #deeplearning #datascience #fair #sampling #samplingbias #python #pytorch #tensorflow #pythonprogramming #learnpython #learnprogramming #engineering #data #algorithms #machinelearningalgorithms #shorts #short
