Gradient Descent

I believe I finally understand what Gradient Descent is! It has a self-regulating feature. Naturally, the slope decreases as it transitions from very steep to less steep and eventually becomes flat (in a parabolic, ideal, simplified environment). Additionally, when it moves downhill, the slope is negative, indicating an uphill direction. We leverage these two characteristics to make well-informed guesses for testing different values of w to help us find the minimum cost.