How to train and test a neural network using scikit-learn and Keras in Jupyter Notebook.
I compare training a neural network in Keras with scikit-learn (MLPRegressor) in Jupyter Notebook.
I show how to train them in both packages and discuss important findings and differences in using them.
0:00 Introduction
0:25 What is scikit-learn?
1:00 sklearn.neural_network.MLPRegressor introduction
1:38 What is Keras?
2:45 Specifications of our neural network for this example
5:24 Dataset used in this example; from UCI Machine Learning repository
6:25 Import modules from sklearn.
7:55 Look at the dataset quickly
10:42 Instantiate the MLPRegressor module
10:59 Deep dive into the MLPRegressor module; Look at the arguments.
15:50 Train/fit MLPRegressor and compute errors
17:35 Keras neural network training begins here; import modules for training.
25:55 Conclusions from this exercise
32:50 Check out my other videos on forward and back propagation.
Пікірлер: 8
Great video!
I have no idea why this video did not explode. You cover everything in good detail, don't really waste time and always spot on! Thanks man!
Simple, easy good for beginner,. I have shared this video to my student for research proposed. Million thx 👍👍👍
Nice. Ty
❤️❤️
Hi, very good tutorial, thanks for sharing. However there's one important detail you left out, which is to normalize/scale the dataset.
@bevansmithdatascience9580
2 жыл бұрын
Thanks, you are right. Didn't want to focus on that in this video but preprocessing the data is vital to performing successful ML project
great video, thank you so much!