2. Regression

I'm sorry, your answer is not correct.

• It tries many different regression lines and converges on a solution with the highest loss value is incorrect. The line with the highest loss value would be the worst solution, not the best.
• It tries many different regression lines and converges on a solution with a loss of zero is incorrect. Linear regression never reaches a loss of zero because each prediction is always slightly wrong. A loss of zero indicates that there is something wrong with the training process.
• It tries many different regression lines and converges on a solution with the lowest loss value is correct
• It calculates the best possible line instantly using advanced differential calculus is incorrect. Linear regression does use differential calculus, but the calculation is iterative, not instantaneously.