From the course: Supervised Learning Essential Training
Unlock the full course today
Join today to access over 22,400 courses taught by industry experts or purchase this course individually.
How neural networks find patterns in data - Python Tutorial
From the course: Supervised Learning Essential Training
How neural networks find patterns in data
- We're going to look at what happens in a neural network as it digests data. Let's take a multi-layer network that takes in two inputs; the weight and height of a cow. First, our weights are set randomly. So despite the fact that we'll get integers for the two inputs, they'll start off being multiplied by random weights, and then those weights will be adjusted to reduce our prediction error. The activation function in our node processes the input value and randomly sets weights to produce an output. Since we use neural networks for supervised learning or cases where we have labeled ground truth data, we compare our models output with our actual value for a particular cow. Then, we'll calculate how far our model's prediction was and update the weights to try and reduce our error. The first step we take to multiply our inputs by our weights is called forward propagation. The output values are usually a probability of an…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.