CSC 4320/6320 Operating Systems will move from CS 200 to a New

advertisement
Assignment #3
CSc 4740/6740 Data Mining
(email your solution (.doc file) to yzhang@cs.gsu.edu,
email subject is Data Mining Assignment #3 )
Due 3/14/2011 (Monday)
Note: Every student must do independent work.
For undergraduate students:
1. (100 points) Please use the given BP (back-propagation learning
algorithm) program
(Please download a source code file from
http://www.cs.gsu.edu/~cscyqz/courses/ci/bp/bp.zip, then modify , and
finally compile it.
you may also download bpPC.exe
http://www.cs.gsu.edu/~cscyqz/courses/ci/bpPC.exe
or bpLinux.exe http://www.cs.gsu.edu/~cscyqz/courses/ci/bpLinux.exe.).
Please personally select a training data set of 100 values from one
domain from http://www.forecasts.org/data/. Then use the training data to
train a neural network, finally use the trained neural network to predict 10
values by using 10 new testing data from the domain from
http://www.forecasts.org/data/.. You need to normalize all input/output
values in [0, 1]. Please use two different Figures to show training results and
testing results (showing both values generated by neural networks and real
values from the Website), and analyze the results. You must show basic
steps clearly (show all data sets and processing steps).
How to Use the Neural Network Software (Back propagation algorithm)
Please download a source code file from
http://www.cs.gsu.edu/~cscyqz/courses/ci/bp/bp.zip
you may also download bpPC.exe
http://www.cs.gsu.edu/~cscyqz/courses/ci/bpPC.exe
or bpLinux.exe http://www.cs.gsu.edu/~cscyqz/courses/ci/bpLinux.exe.
(1) Train a neural network command:
bpPC LEARN olddata.dat weight.net 0.45 0.9 0.001 100 2 4 1
Or
bpLinux LEARN olddata.dat weight.net 0.45 0.9 0.001 100 2 4 1,
where, bpPC, bpLinux : program name
LEARN: to train a neural network
olddata.dat: past stock data file
weight.net: neural network parameter file
0.45, 0.9: parameters (you don’t need to change them)
0.001: training error, you may change it
100: training cycles
2: number of inputs (input neurons)
4: number of hidden neurons
1: number of outputs (output neurons)
For example: The XOR problem is solved by the BP learning algorithm.
bpPC LEARN xor.dat weight.net 0.45 0.9 0.001 100 2 4 1
xor.dat is
1 0.1 0.1 0.1
2 0.1 0.9 0.9
3 0.9 0.1 0.9
4 0.9 0.9 0.1
The 1st column is sequential number, then 2 input columns and one
output column.
(2) Use the trained neural network to predict future values
bpPC RUN newdata.dat weight.net,
or
bpLinux RUN newdata.dat weight.net.
where,
RUN: run the trained neural network to generate new outputs based on
newdata.dat
For graduate students:
1. (100 points) Please use the given BP (back-propagation learning
algorithm) program
(Please download a source code file from
http://www.cs.gsu.edu/~cscyqz/courses/ci/bp/bp.zip, then modify , and
finally compile it.
you may also download bpPC.exe
http://www.cs.gsu.edu/~cscyqz/courses/ci/bpPC.exe
or bpLinux.exe http://www.cs.gsu.edu/~cscyqz/courses/ci/bpLinux.exe.).
Please personally select two different data sets of 200 values from
two different domains from http://www.forecasts.org/data/. Make a
training data set with two inputs from domain 1, two inputs from domain 2,
and two outputs from domains 1 and 2 (i.e., 4 input-2 output training data
set), then train a neural network, finally use the trained neural network to
predict 10 values by using 10 new testing data from from the two different
domains from http://www.forecasts.org/data/. You need to normalize all
input/output values in [0, 1]. Please use two different Figures to show
training results and testing results (showing both values generated by neural
networks and real values from the Website), and analyze the results. You
must show basic steps clearly (show all data sets and processing steps).
How to Use the Neural Network Software (Back propagation algorithm)
Please download a source code file from
http://www.cs.gsu.edu/~cscyqz/courses/ci/bp/bp.zip
you may also download bpPC.exe
http://www.cs.gsu.edu/~cscyqz/courses/ci/bpPC.exe
or bpLinux.exe http://www.cs.gsu.edu/~cscyqz/courses/ci/bpLinux.exe.
(1) Train a neural network command:
bpPC LEARN olddata.dat weight.net 0.45 0.9 0.001 100 2 4 1
Or
bpLinux LEARN olddata.dat weight.net 0.45 0.9 0.001 100 2 4 1,
where, bpPC, bpLinux : program name
LEARN: to train a neural network
olddata.dat: past stock data file
weight.net: neural network parameter file
0.45, 0.9: parameters (you don’t need to change them)
0.001: training error, you may change it
100: training cycles
2: number of inputs (input neurons)
4: number of hidden neurons
1: number of outputs (output neurons)
For example: The XOR problem is solved by the BP learning algorithm.
bpPC LEARN xor.dat weight.net 0.45 0.9 0.001 100 2 4 1
xor.dat is
1 0.1 0.1 0.1
2 0.1 0.9 0.9
3 0.9 0.1 0.9
4 0.9 0.9 0.1
The 1st column is sequential number, then 2 input columns and one
output column.
(2) Use the trained neural network to predict future values
bpPC RUN newdata.dat weight.net,
or
bpLinux RUN newdata.dat weight.net.
where,
RUN: run the trained neural network to generate new outputs based on
newdata.dat
Download