I'm studying a Master's Degree in Artificial Intelligence and I need to learn how to use the Java Neural Network Simulator, JavaNNS, program.
In one practice I have to build a neural network to use backpropagation on it.
I have created a neural network with one input layer with 12 nodes, one hidden layer with 6 nodes and one output layer with 1 node.
I'm using Kaggle's titanic competition data with this format following Dataquest course Getting Started with Kaggle for Titanic competition:
Pclass_1,Pclass_2,Pclass_3,Sex_female,Sex_male,Age_categories_Missing,Age_categories_Infant,Age_categories_Child,Age_categories_Teenager,Age_categories_Young Adult,Age_categories_Adult,Age_categories_Senior,Survived 0,0,1,0,1,0,0,0,0,1,0,0,0 1,0,0,1,0,0,0,0,0,0,1,0,1 0,0,1,1,0,0,0,0,0,1,0,0,1 1,0,0,1,0,0,0,0,0,1,0,0,1 0,0,1,0,1,0,0,0,0,1,0,0,0 0,0,1,0,1,1,0,0,0,0,0,0,0 1,0,0,0,1,0,0,0,0,0,1,0,0 0,0,1,0,1,0,1,0,0,0,0,0,0
But they preprocess the data to use it with linear regression and I don't know if I can use these data with backpropagation
I think something is wrong because when I run backpropagation in JavaNNS I get these data:
opened at: Sat Feb 17 17:29:40 CET 2018 Step 200 MSE: 0.5381023044692738 validation: 0.11675894327003862 Step 400 MSE: 0.5372328944712378 validation: 0.11700781497209432 Step 600 MSE: 0.5370386219557437 validation: 0.11691717861750939 Step 800 MSE: 0.5370348711919518 validation: 0.11696104763606407 Step 1000 MSE: 0.5369724294992798 validation: 0.11697568840154722 Step 1200 MSE: 0.5369697016710676 validation: 0.11665485957481342 Step 1400 MSE: 0.5370053339270906 validation: 0.11684215268609244 Step 1600 MSE: 0.5370121961199371 validation: 0.11670833992558485 Step 1800 MSE: 0.5370200812483633 validation: 0.11673550099633925 Step 2000 MSE: 0.5367923502149529 validation: 0.11675956129361797
Nothing changes, it is like it doesn't learn anything.
How many hidden layers does the network have with how many nodes on each hidden layer?
Maybe the problem is that the data have been prepared to be used in Linear regression and I using it with Backpropagation.
I have only created the neural network, I haven't implemented the backpropagation algorithm because it is already implemented in JavaNNS.