From the course: Neural Networks and Convolutional Neural Networks Essential Training
Join today to access over 25,300 courses taught by industry experts.
Activation functions
From the course: Neural Networks and Convolutional Neural Networks Essential Training
Activation functions
We've built our neural network architecture, but there's still one crucial missing piece. Remember our XOR problem? Well, we solved it by using multiple layers, but there's a catch. If every neuron just adds up its inputs and passes them along, we're still stuck with straight line thinking. It's like having a team of people who all have exactly the same personality. They'll all react to situations in exactly the same predictable way. Now, what if we could give each neuron a unique personality, a specific way of reacting to information? And that's exactly what activation functions do. They're like personality types that determine how a neuron responds when it receives information. Let me introduce you to the different personality types in our neural network family. The step function, this is the all or nothing personality. Now like that friend who either loves something completely or hates it with no middle ground, if the input is above a certain threshold, the output is 1. Below that…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.