untitled.py

# -*- coding: utf-8 -*-
"""Untitled.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1gcDwD4fUxTVGnRbEmZoIYGDhCeTNYhrt
"""

import pandas as pd
df = pd.read_csv('sample_data/ramen-ratings.csv')

from keras.layers import Input, LSTM, Bidirectional, SpatialDropout1D, Dropout, Flatten, Dense, Embedding, BatchNormalization
from keras.models import Model
from keras.callbacks import EarlyStopping
from keras.preprocessing.text import Tokenizer, text_to_word_sequence
from keras.preprocessing.sequence import pad_sequences
from keras.utils import to_categorical
import nltk, os, re, string
from nltk.corpus import stopwords

Style = pd.get_dummies(df.Style)
df_baru = pd.concat([df, Style], axis=1)
df_baru = df_baru.drop(columns='Style')

df_baru

news = df_baru['Variety'].values
label = df_baru[['Bar',	'Bowl',	'Box',	'Can',	'Cup',	'Pack',	'Tray']].values

from sklearn.model_selection import train_test_split
news_train, news_test, label_train, label_test = train_test_split(news, label, test_size=0.2)

from tensorflow.keras.preprocessing.text import Tokenizer
from tensorflow.keras.preprocessing.sequence import pad_sequences
 
tokenizer = Tokenizer(num_words=5000, oov_token='zn')
tokenizer.fit_on_texts(news_train) 
 
sekuens_latih = tokenizer.texts_to_sequences(news_train)
sekuens_test = tokenizer.texts_to_sequences(news_test)
 
padded_latih = pad_sequences(sekuens_latih) 
padded_test = pad_sequences(sekuens_test)

import tensorflow as tf
from tensorflow.python.keras import regularizers
model = tf.keras.Sequential([
    tf.keras.layers.Embedding(input_dim=5000, output_dim=16),
    tf.keras.layers.LSTM(64),
    tf.keras.layers.Dense(128,activation = 'relu', kernel_regularizer=regularizers.l2(0.001)),
    tf.keras.layers.Dropout(0.5),
    tf.keras.layers.Dense(7, activation='softmax')
])
model.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy'])
model.summary()

class myCallback(tf.keras.callbacks.Callback):
  def on_epoch_end(self, epoch, logs={}):
    if(logs.get('accuracy')>0.75 and logs.get('val_accuracy')>0.75):
      print("\nAkurasi pada training set dan validation set telah mencapai >70%!")
      self.model.stop_training = True
callbacks = myCallback()

num_epochs = 50
history = model.fit(padded_latih, label_train, epochs=num_epochs, 
                    validation_data=(padded_test, label_test), verbose=2, callbacks=[callbacks])

import matplotlib.pyplot as plt
plt.plot(history.history['accuracy'])
plt.plot(history.history['val_accuracy'])
plt.title('Akurasi Model')
plt.ylabel('accuracy')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
plt.show()

plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('Loss Model')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
plt.show()