You are on page 1of 6

Nova Southeastern University College of Engineering and Computing

MMIS 643 - Written Assignment #3B

Consider the Boston Housing Data file (The schema of the data file is given on page 33 in Table 2.2 of the
textbook. )(15 points)

a. Study the Neural Networks Prediction example from the URL:


http://www.solver.com/xlminer/help/neural-networks-classification-intro, and following the example
step by step (using the manual neural network classification example)

b. Using XLMINER’s neural network routine under predict menu to fit a model using XLMINER
default values for neural network parameters by using the predictors such as CRIM, ZN, INDUS,
CHAS, NOX, RM, AGE, DIS, RAD, TAX, PTRATIO, B, LSTAT to predict the value of the outcome variable
MEDV.

i. Record the RMS errors for the training data and the validation data, and observe the lift
charts for repeating the process, changing the number of epochs to 300, 3000, 10,000,
20,000.

Training Validation
Data RMS Data RMS
# of Epochs Error Error
4.56717175
300 6 4.22927402
3.36923221
3000 9 3.897988691
3.27186568
10000 1 3.722164633
3.24666831
20000 3 3.666689902

This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/
This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/
This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/
This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/
Training Validation
# of Epochs Data Data
4.56717175
300 6 4.22927402
3.36923221
3000 9 3.897988691
3.27186568
10000 1 3.722164633
3.24666831
20000 3 3.666689902

ii. What happens to RMS error for the training data set as the number of epochs increases?

When training the model, we observed decreasing root mean square error (RMS error) over an increasing
number of epochs. After 300 epochs, the training error was higher than validation error. But with continual
training, the training RMS error lowered substantially with additional epochs - dropping below the validation
error.

Key points:
1. More epochs drove down RMS error
2. Training error higher at 300 epochs
3. Continued epochs reduced training error
4. Training RMS error fell lower than validation error

iii. What happens to RMS error for the validation data set as the number of epochs increases?

As the epochs increased the RMS Error decreased.

iv. Comments on the appropriate number of epochs for the model.


Note: (Please use the Prediction Option of the Neural Network in order to get RMS error)
This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/
20,000 is the appropriate number of epochs since this is when the RMS error is at its lowest point for both
the training data and validation data

This study source was downloaded by 100000769334500 from CourseHero.com on 11-19-2023 15:49:46 GMT -06:00

https://www.coursehero.com/file/22554616/MMIS-643-Written-Assignment-3B/

You might also like