# Data prepared to linear regression. Can I use it with backpropagation?

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


If you want see the same data better in an Spreadsheet:

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.

• I'm pretty sure you are implementing the backprop wrong...It is notoriously difficult to implement.. Check this out stackoverflow.com/questions/41872353/… – DuttaA Feb 17 '18 at 17:35
• Or you can head over to Andrew ngs lecture on testing backprop..Just do a quick Google search – DuttaA Feb 17 '18 at 17:36
• @DuttaA I haven't implemented the backpropagation algorithm because it is already implemented in JavaNNS. – VansFannel Feb 17 '18 at 19:07
• Have you preprocessed your data properly? – Daniel Feb 17 '18 at 23:39
• @Daniel I have updated my question with more details about how I preprocess the data. Thanks. – VansFannel Feb 18 '18 at 7:32