how to handle multi-classification using one-vs-rest method?

[simsong]

From @jm-huang on April 19, 2016 9:40

I am a little confusing while using this package for multi-classification. can anyone tell me how to do it ? Thanks.

what i had try:

train_labels=[[1,2], [2], [3]]
train_datas = [[1,1,0], [1,2,2], [1,1,1]]
prob = problem(train_labels, train_datas)
param = parameter('-s 0')
model = train(prob, param)

but it arise some errors:
Traceback (most recent call last):
File "C:\Users\Jiaming\Dropbox\Internship in ADSC\DeepWalk\experiments\classifier.py", line 69, in process
prob = problem(train_labels, train_datas)
File "C:\Users\Jiaming\Anaconda2\lib\site-packages\liblinear-210-py2.7.egg\liblinear\liblinear.py", line 107, in init
for i, yi in enumerate(y): self.y[i] = y[i]
TypeError: a float is required

Copied from original issue: cjlin1#21

[simsong]

From @cjlin1 on April 19, 2016 10:6

you don't need to handle it. Liblinear directly supports 1-vs-rest

Billy writes:

I am a little confusing while using this package for
multi-classification. can anyone tell me how to do it ? Thanks.

what i had try:

train_labels=[[1,2], [2], [3]]
train_datas = [[1,1,0], [1,2,2], [1,1,1]]
prob = problem(train_labels, train_datas)
param = parameter('-s 0')
model = train(prob, param)

but it arise some errors:
Traceback (most recent call last):
File "C:\Users\Jiaming\Dropbox\Internship in ADSC\DeepWalk
experiments\classifier.py", line 69, in process
prob = problem(train_labels, train_datas)
File "C:\Users\Jiaming\Anaconda2\lib\site-packages
liblinear-210-py2.7.egg\liblinear\liblinear.py", line 107, in
init
for i, yi in enumerate(y): self.y[i] = y[i]
TypeError: a float is required

—
You are receiving this because you are subscribed to this
thread.
Reply to this email directly or view it on GitHub*

[simsong]

From @jm-huang on April 19, 2016 10:16

@cjlin1 I think that I am using the wrong data format, can you show me the format of the train data file ? or the correct format of "problem"?
Thanks

[simsong]

From @jm-huang on April 19, 2016 10:17

@cjlin1 That's how can i using train() method in python to do that?

[simsong]

From @jm-huang on April 20, 2016 1:18

can you kindly tell me how can i set the format of "train_labels" with multi-class ? what i did make wrong results. Thanks very much.
train_labels=[[1,2], [2], [3]]
train_datas = [[1,1,0], [1,2,2], [1,1,1]]
prob = problem(train_labels, train_datas)
param = parameter('-s 0')
model = train(prob, param)

[simsong]

From @jm-huang on April 20, 2016 1:30

I using this format: the first part is the labels, and the second part is the features.
train.txt:
1,2,3 3 4 5
2 4 5 5
3 3 4 5

train train.txt
and it output:
Wrong input format at line 1
I just don't know how to handle this error.

[simsong]

From @rofuyu on March 7, 2017 20:37

You have a multi-label dataset (more than 1 positive labels for each instance) instead of a mulit-class dataset. The current liblinear only supports multi-class classification. If the number of labels in your case is small, you can re-label them using the index to the power set: e.g.,
{1} -> 1
{2} -> 2
{3} -> 3
{1,2} -> 4
{1,3} -> 5
{2,3} -> 6
{1,2,3} ->7.