In the same figure, if we draw a tangent at the green point, we know that if we are moving upwards, we are moving away from the minima and vice versa. We will talk about this in more detail in the latter part of the article. So, if we can compute this tangent line, we might compute the desired direction to reach the minima. The slope is described by drawing a tangent line to the graph at the point. A derivative is a term that comes from calculus and is calculated as the slope of the graph at a particular point. Gradient Descent Algorithm helps us to make these decisions efficiently and effectively with the use of derivatives. which way to go and how big a step to take. If you decide which way to go, you might take a bigger step or a little step to reach your destination.Įssentially, there are two things that you should know to reach the minima, i.e.In a Cartesian coordinate system, this is an equation for a parabola and can be graphically represented as : If you want your logo to have a container, slogan, or all of the above, adding a gradient will make the design too busy and hard to digest. If you want a wordmark (text-only) logo, a gradient design probably isn’t for you. If we look carefully, our Cost function is of the form Y = X². Instead, gradients are usually applied to the symbol or monogram elements in a logo.
Since we want the lowest error value, we want those‘ m’ and ‘ b’ values that give the smallest possible error. Or you can start with our AI logo maker by entering your brand name and let our logo engine help you find the perfect design.
Click on the symbol you like to start making your logo. This is because a lower error between the actual and predicted values signifies that the algorithm has done an excellent job learning. Enter a keyword to search our extensive logo symbol library. The goal of any Machine Learning Algorithm is to minimize the Cost Function. Also, the squared differences increase the error distance, thus, making the bad predictions more pronounced than the good ones. Indeed, to find that line we need to compute the first derivative of the Cost function, and it is much harder to compute the derivative of absolute values than squared values. Why do we take the squared differences and simply not the absolute differences? Because the squared differences make it easier to derive a regression line.