Weka Tutorial 23: Classification 101 using API (Classification)

[@suckyboy]

Thanks for the great work : )
Keep making more awesome weka API tutorials!

[@Machimachina]

This stuff on the API was very helpful. I'd love to see some more stuff on it, diving into doing stuff that isn't possible with the WEKA GUI.

[@onecelledboy]

is there a way to determine how closely your training data fits the test data as it is classifying the training data?

[@zakiyaafni784]

Thanks a lot for the videos :)
I wanna ask why does the number of my labeled-data's result is not same with the number of my test data.. it is decrease and I think it didn't all of my test data was labeled..

[@WingsofDestiny123]

This was incredibly helpful. Thank you for uploading this video.

[@rushdishams]

Thanks a lot.

[@filmyduck7887]

Thanks for the great video! I am working on Android Studio. If I want my phone to have access to the .arff files, where should I put them? What would be their file paths?

[@darmayoga9776]

best video that im looking for

[@profmatheusfranco]

Hi, I can download the java code of this example in some place, or you can send me? Best.

[@amarrocks1231]

Sir what is the point of classifying if you use the same data as the training set(though unlabeled) in the test set? It just relabels the apparently unknown data with correct labels, but what if I wanted to test it on an entirely different data?

[@rushdishams]

Amar Yadav just for an example. I was lazy to create a new dataset.

[@amarrocks1231]

+Rushdi Shams Actually Sir I meant something else, until yesterday I thought I cannot use the trained classifiers to classify new instances read from new testing.arff file from within WEKA Explorer's classify tab. Upon trying to do so, I used to get no. of total instances as 0 in the evaluation summary. After two frustrating days, I found out that my Mac's in-built text editor tampered the integrity of the copy of training.arff that I used to make testing.arff, i.e. the testing.arff got corrupted the moment I opened and edited it with TextEdit. I concluded this because the WEKA explorer gave weird error on opening the testing.arff file under pre-processing tab. I switched to Sublime Text and no problems since then. I thought I'll mention this so that no one else has to spend a day banging their head. But thanks for the reply sir!

[@amyesmeralda87]

Thanks for your Videos.
-Amy E.

[@anjalisaini3299]

Hii sir
can hybrid svm + Random forest in weka.

[@charan6912]

Hi Rushdi. Thanks for the code, it helped me a lot. Could you please let me know how to classify two classes for the same arff file simulatneously.

[@rushdishams]

You need to create two objects. One will be, say, with the dataset with one classifier and the other one will be, say, with the dataset with a different classifier.

[@hiteshsethi8516]

can you please tell how can i change kernel type of LibSvm in weka api. It says use setKernelType(); i have tried that but it takes selectedTag as input. Am not able to get that.

[@rushdishams]

Take a look at the libsvm documentation. It might help you.

[@manshan9751]

Hi, thanks a lot for this video. but how to do the same in explorer

[@rolanflores3125]

Thanks for this video it help me a lot
-Rolan

[@elvieolapaniamaro8033]

Thanks for your videos ❣️
Elvie A.

[@IgbeTobore]

Thanks! just what I needed

[@alialQatif08]

what was the main class you chose to run it?

[@rushdishams]

Hi. It was just one class I named Classify

[@marjoriesabra9778]

Thank you 4 diz tutorial.
-marjorie

[@IsaacLopes90]

I use the same code as you do, but with different 'train' and 'test' data files. In result an empty 'labeled.arff' file is created. Can you tell What is the problem?

[@luckyluke7684]

+Isaac Lopes I think you are missing these line writer.newLine();writer.flush();writer.close();

[@filmyduck7887]

@@luckyluke7684 Where do you put the writer.newLine(); Before or after the writer.write(labeled.toString());?
Right now, I have it as:
writer.newLine();
writer.write(labeled.toString());
writer.flush();
writer.close();

[@sonnyberja9106]

Thank you ❤️
-Sonny

[@ricamaemitra1104]

Thanks a lot for this video tutorial.
-RICA MAE

[@brelopes6426]

Hello sir, I followed your example but I got a blank result, any help?

[@rushdishams]

There can be many reasons behind this. I need to take a brief look at the dataset.

[@luckyluke7684]

+bre Lopes I think you are missing these line writer.newLine();writer.flush();writer.close();

[@VishalGangwar081]

Can you do the same thing for android? I have done the same thing, it works fine in java but didn't work in android

[@rushdishams]

I am sorry that I cannot help you on this as I don't work with androids.

[@luckyluke7684]

+Vishal Gangwar I think you can do the same in android because the core of android is java, so you just code as this and save to prefference

[@pravinkharat2990]

I followed the steps in your tutorial. But I got result file as blank. Can you suggest where am I going wrong ?

[@rushdishams]

Something wrong with the classification I guess?

[@luckyluke7684]

+Pravin Kharat I think you are missing these line writer.newLine();writer.flush();writer.close();

[@filmyduck7887]

@@luckyluke7684 That doesn't seem to work. I get an error: error: illegal character: '\ufeff'

[@annabelrubia7153]

Thank you!
-Annabel

[@jaleel3853]

i have problem with weka when use arabic language

[@naglaagaze6691]

thanks for your videos

[@0xkellan]

hi sir, can i get all codes from this video?

[@anglikai9517]

Thank you !

[@reyannarcilla7141]

thankyou for this.
-reyann

[@angelicamariano5139]

Thanks....
-Angelica

[@lezeildacuba5001]

Thankyou

[@hayfaazibi8779]

thank you so much :)

[@ma.ninaaragdon5044]

Thank you.
-Niña Aragdon

[@rubelynalarcon9492]

Rubelyn ❤
Done Thank You Po .