Artificial Neural Nets
Machine Learning works on the basis of Artificial Neural Nets (ANNs). These neural nets are meant to resemble how the human brain processes information. The neural net is broken up into three sections, the inputs, the hidden layer, and the outputs. From the Inputs, those values are given a weight (significance) and then determined a bias (how trustworthy the data is). From there it enters the Hidden layer where it gets put through an activation function such as the sigmoid function of the tanh function. This changes the input data into values easier for the program to interpret and stops the program from limiting the data to a linear curve. The data leaving the hidden layer is then reassigned a weight and bias and sent to the outputs. These outputs are then evaluated on how far off they were using gradient descent and the weights are adjusted to better optimize the outputs the next time it is run through.