-
Notifications
You must be signed in to change notification settings - Fork 0
/
manual.txt
21 lines (13 loc) · 947 Bytes
/
manual.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Manual for running code:
1) Run main.py
2) Type 1 to run the linear perceptron. We changed the number of iterations to 100. However, in our final report we did 1000 and 10,000. This is so when the grader runs it will complete in around 60 seconds (at worst). To modify this, change the iteration number in linear_perceptron.py (Line 43).
3) Type 2 to run the radial basis kernel support vector machine. This should run in around 60 seconds (at worst).
4) Type 3 for RBF SVM K fold Cross Validation and specify the number of folds (We recommend 10), the slack value, and whether or not you want to shuffle the data beforehand. This should run in 30 seconds assuming 10 folds, shuffling, and C=100.
5) Type 4 to see all the graphs. There should be 17 figures in total.
External libraries and tools:
mandatory:
pip install PrettyTable
other installations if needed:
pip install -U scikit-learn
pip install pandas
pip install matplotlib