adult_dataset_svm_libsvm

Data preprocessing

• Basic way： only cleaning data (mostly deleting missing value), digitalization and discretization.
• Handling dummy variables besides process above

SVM

• Using svm in python package sklearn (without adjustment parameters)
• Ten-fold cross validation

libsvm

• Using libsvm, a svm library developed by Professor Chih-Jen Lin of Taiwan in 2001
• The training set and test set are divided at a ratio of 2:1. 'a9a' is training set while 'a9a.t' is test set. This two sets were in the format of libsvm, which came from https://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary.html#a9a

More

• adult_libsvm_self.py is a file that I try converting the data I preprocessed to libsvm format then using libsvm on it, but a little problems occured. Still improving...

Thanks

appreciation for work on https://blog.csdn.net/gengkui9897/article/details/83049036 and https://blog.csdn.net/dianqijiaojianshuo/article/details/80620811