sid321axn

model_lstm.compile(loss=tf.keras.metrics.mean_squared_error,
              metrics=[tf.keras.metrics.RootMeanSquaredError(name='rmse')], optimizer='adam')
early_stop = EarlyStopping(monitor='loss', patience=5, verbose=1)
history_model_lstm = model_lstm.fit(X_tr_t, y_train, epochs=100, batch_size=1, verbose=1, shuffle=False, callbacks=[early_stop])

sid321axn

from keras.layers import LSTM
K.clear_session()
model_lstm = Sequential()
model_lstm.add(LSTM(50, input_shape=(1, X_train.shape[1]), activation='relu', kernel_initializer='lecun_uniform', return_sequences=False))
model_lstm.add(Dense(1))
model_lstm.summary()

sid321axn

X_tr_t = X_train.reshape(X_train.shape[0], 1, X_train.shape[1])
X_tst_t = X_test.reshape(X_test.shape[0], 1, X_test.shape[1])

sid321axn

model.compile(loss=tf.keras.metrics.mean_squared_error, metrics=[tf.keras.metrics.RootMeanSquaredError(name='rmse')], optimizer='adam')

early_stop = EarlyStopping(monitor='loss', patience=2, verbose=1)

history = model.fit(X_train, y_train, epochs=200, batch_size=1, verbose=1, callbacks=[early_stop], shuffle=False)

sid321axn

K.clear_session()

model = Sequential()
model.add(Dense(12, input_dim=1, activation='relu'))
model.add(Dense(1))
model.summary()

sid321axn

X_train = train_sc[:-1]
y_train = train_sc[1:]

X_test = test_sc[:-1]
y_test = test_sc[1:]

sid321axn

from sklearn.preprocessing import StandardScaler

sc = StandardScaler()
train_sc = sc.fit_transform(train)
test_sc = sc.transform(test)

sid321axn

#setting the split date
split_date = pd.Timestamp('26-08-2018')

# creating training dataframe
train = df.loc[:split_date]
# creating test dataframe
test = df.loc[split_date:]

#plotting train test dataframe as aline plot
ax = train.plot(kind='line',figsize=(12,8))

sid321axn

df = data_frame[["Price"]]

#plotting dataset to visualize the pattern of prices over the years

df.plot(kind='line',figsize=(12,7))

sid321axn

# sorting the dataset in ascending order of date
data_frame = ind_exchange_data.sort_index(axis=1 ,ascending=True)
data_frame = data_frame.iloc[::-1]
data_frame.head()