How an AI learns, in 3 small steps · part 2 of 3 · Error, squared error and the loss (MSE)
How does an AI know it’s wrong?
For each student, the error: the gap between the prediction and the real mark. Square each error, take the mean, and the AI has one number that says how wrong it is: the loss, here the mean squared error.
What you turn: Weight and bias
Next in the Cogito playlist
The reel