Python Deep Learning Tips - How to Buil a Feed Forward Network in Tensorflow with the Sequential API

Feed Forward Neural Network

In Python using Tensorflow design a Feed Forward neural network to predict a regression problem with 40 features. To stick to the style of a Feed Forward architecture we continue with 40 neurons down our Dense or fully connected layers. Although adding in more Dense layers will help at first the comes a point we more Dense layers won't help with the problem any further.

A good way to look at this is your dataset has only a certain number of higher-level features discoverable by the network and adding more layers when there are no more higher features to find won't add value. How many layers? Experimentation is the solution to finding the perfect amount for your data set.

Data Science Teacher Brandyn YouTube Channel

One on one time with Data Science Teacher Brandyn

Follow Data Science Teacher Brandyn

On Facebook

On Linkedin

dataGroups:

Showcase your DataArt on facebook

Showcase your DataArt on linkedin

Python data analysis group, share your analysis on facebook

Python data analysis on linkedin

Machine learning in sklearn group

Join the deep learning with tensorflow facebook group

Join the deep learning with tensorflow on linkedin

Feed Forward Deep Learning Model in Tensorflow Sequential Api

Feed Forward Predict better than Perceptron

More Dense Layers will only make a difference to a certain point

With 3 more Dense we see only a little improvement in the predictions in our fully connected Fee Forward Network