From the course: Building Deep Learning Applications with Keras
Join today to access over 24,400 courses taught by industry experts.
The Train-Test-Evaluate cycle
From the course: Building Deep Learning Applications with Keras
The Train-Test-Evaluate cycle
- [Instructor] "Neural Networks in Keras: One Simple Path" Well, this simple path has four steps: model selection, training phase, testing phase, and evaluation phase. Let's go over them one by one. Supervised learning is just like guiding a student. We provide the input and the expected results to the model, and we teach it to transform inputs into the correct outputs. Remember the Keras High School example we had earlier in the course? Well, that student, in this case the model, is eager to learn the patterns of arithmetic. That's our neural network here, and we are the tutors providing the right problems, like input data, and solutions, like expected results. Our student then learns the art of prediction, aiming to ace the final exams, which is the testing place. In the model selection, we start by selecting a neural network algorithm from Keras's very extensive library. Then we divide the data into training and testing. Once we are done with the model, we will show the model some…
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.
