search

EklavyaFCB / EMNIST-Kohonen-SOM

Using Kohonen's algorithm to make a Self Organising Map using the EMNIST database on handwritten alphabets.
36 stars 12 forks source link

The training data, test data and label files were not found in the hosting university server, can they be uploaded in the repository or tell us from where they can be downloaded? #1

Closed nsapkota68 closed 4 years ago

EklavyaFCB commented 4 years ago

I forget exactly how I unwrapped the data, but you can find it here: https://github.com/aurelienduarte/emnist

Then you have to sort it with sort.py and put the correct path in som.py as explained here: https://github.com/EklavyaFCB/EMNIST-Kohonen-SOM#note

nsapkota68 commented 4 years ago

Thank you so much.

Nabin

On Sat, Sep 26, 2020 at 11:38 AM Eklavya Sarkar notifications@github.com wrote:

I forget exactly how I unwrapped the data, but you can find it here: https://github.com/aurelienduarte/emnist

Then you have to sort it with sort.py and put the correct path in som.py: https://github.com/EklavyaFCB/EMNIST-Kohonen-SOM#note

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/EklavyaFCB/EMNIST-Kohonen-SOM/issues/1#issuecomment-699518427, or unsubscribe https://github.com/notifications/unsubscribe-auth/AQOA4W5VJGZBML4HSZCIJDDSHYKJZANCNFSM4R2WEMEQ .

nsapkota68 commented 4 years ago

dataset_byclass.txt

I was able to read the Matlab file, but could not process nested nd array in a file 'dataset_byclass'. I have spent quite a time trying to unwrap it, then sort it, but it seems like it is beyond my skill. Any help is appreciated. I want to use the results in a paper and cite this source. Thank you.

The file attached is the values of the dataset field of dictionary item of the Matlab file dataset_byclass.

EklavyaFCB commented 4 years ago

Alright I'll try to sort this tomorrow morning and give it to you asap, I need to update this repo anyway.

EklavyaFCB commented 4 years ago

I have updated the instructions, and placed the new .csv files on my google drive.

You just have clone again (or pull), place the data from my GDrive to the correct path as explained here, and run som.py.

Have also added a cite part.

Let me know if you have any issues now.

For further reading, you can also refer to my thesis.

nsapkota68 commented 4 years ago

Thank you very much, I could run all the files now. Your thesis is impressive.

EklavyaFCB commented 4 years ago

Great! I look forward to reading your paper too someday !

EklavyaFCB commented 3 years ago

Did you ever write a paper ?