Ever Wonder What Bias Looks Like in Machine Learning?
Connect the dots on steroids.
The term “machine learning” is becoming increasingly popular, but what is less often discussed is that machines don’t learn per se. They are programmed by human beings to process more sophisticated information in more complex ways, which enables the machine to do more without ongoing human intervention. But software is language, remember. And language tends to lend itself bias, whether intentional or unintentional. Here’s a fantastic explanation of what bias looks like when realized through a machine.
h/t FlowingData
