Normalization
Pennsylvania State University - Harrisburg
School of Business Administration
INFSY 566: Data Mining and Knowledge Discovery
Spring 2014
ASSIGNMENT: Neural Networks for classification
Professor Name: Dr. Parag Pendharkar
Submitted by: Deepak Mahajan Date: February 26, 2014
The strategies used to create training and test data sets are random sampling and stratified sampling. Neural network has been used as modelling technique to create bankruptcy predicting model.
Random Sampling
File conversion from excel to statistical file (.sav) using IBM SPSS statistics. Load file, using statistical file option under Sources tab in SPSS Modeler.
Attached the Type node to the file and selected bankruptcy as target.
Attached Partition node, to create testing and training data file having 50 values randomly picked.
Attached the Analysis node, to perform a check on the partitioned data and exported the same as excel file.
Created two files training and testing to perform analysis by picking up randomly selected first 50 values and named it as testing and the same for the training file.
Converted both the files into statistical file to perform the analysis.
Opened the training statistical file, attached the Type node, and selected the target.
Attached Neural network node to the Type node, selected the basics option under the build option tab and by selecting the customize number of units, assigned 3 in the hidden layers1.
Run the Neural network analysis and attached the Analysis node to the golden nugget to check the results. Followed the last two steps and created two different neural network models by assigning 5, 9 and 15 in hidden layer1 respectively. Stopping rules set to 15 minutes per model, used default values in ensembles, overfit prevention set to 30%.
Lastly, attached the Analysis node to each of the golden nugget retrieved, to check the results for the neural network Model 1:
School of Business Administration
INFSY 566: Data Mining and Knowledge Discovery
Spring 2014
ASSIGNMENT: Neural Networks for classification
Professor Name: Dr. Parag Pendharkar
Submitted by: Deepak Mahajan Date: February 26, 2014
The strategies used to create training and test data sets are random sampling and stratified sampling. Neural network has been used as modelling technique to create bankruptcy predicting model.
Random Sampling
File conversion from excel to statistical file (.sav) using IBM SPSS statistics. Load file, using statistical file option under Sources tab in SPSS Modeler.
Attached the Type node to the file and selected bankruptcy as target.
Attached Partition node, to create testing and training data file having 50 values randomly picked.
Attached the Analysis node, to perform a check on the partitioned data and exported the same as excel file.
Created two files training and testing to perform analysis by picking up randomly selected first 50 values and named it as testing and the same for the training file.
Converted both the files into statistical file to perform the analysis.
Opened the training statistical file, attached the Type node, and selected the target.
Attached Neural network node to the Type node, selected the basics option under the build option tab and by selecting the customize number of units, assigned 3 in the hidden layers1.
Run the Neural network analysis and attached the Analysis node to the golden nugget to check the results. Followed the last two steps and created two different neural network models by assigning 5, 9 and 15 in hidden layer1 respectively. Stopping rules set to 15 minutes per model, used default values in ensembles, overfit prevention set to 30%.
Lastly, attached the Analysis node to each of the golden nugget retrieved, to check the results for the neural network Model 1: