comfy-rec / rnn

pos tagging, sms, rnn, lstm, google colab, pytorch, torchvision, mnist, billboard crawling, selenium, beautifulsoup

rnn(recurrent neural network)

requirements

google colab
pytorch
torchvision pandas numpy selenium beautifulsoup openpyxl

ex.py36.ipynb

pytorch tutorial, mnist data load

py36.형태소분석+품사태깅.ipynb

konlpy package library
Okt, Kkma, Hannanum, Komoran, Twitter, Mecab

py36.RNN.mnist.ipynb

datasets

mnist(28x28 images), 10 calsses

model lstm linear

loss CEL cross entropy loss

optimizer adam

0. Data 전처리.sms.p36.ipynb

data load column word length duplicate shuffle train test split save

1. RNN.sms.p36.ipynb

datasets

sms data

model lstm linear logsoftmax

loss NLLLoss

optimizer adam

billboard_crawling.ipynb

selenium, chromedriver, beautifulsoup, openpyxl
billboard chart crawling
song_list, artist_list
remove '\n'
save 'billdata.tsv'