[ML] Multiple Input Multi-step Output

GITHUB

https://github.com/yiwonjae/Project_Lotto/blob/master/Book_001/p113.py

 

0. 목표

input 多, output 多step

 

1. DATA

in_seq1 = np.asarray([10, 20, 30, 40, 50, 60, 70, 80, 90])
in_seq2 = np.asarray([15, 25, 35, 45, 55, 65, 75, 85, 95])
out_seq = np.asarray([in_seq1[i]+in_seq2[i] for i in range(len(in_seq1))])

in_seq1:ndarray = in_seq1.reshape((len(in_seq1), 1))
in_seq2:ndarray = in_seq2.reshape((len(in_seq2), 1))
out_seq:ndarray = out_seq.reshape((len(out_seq), 1))

dataset = np.hstack((in_seq1, in_seq2, out_seq))

 

2. DATA 정제

X

from numpy import ndarray
import numpy as np


def split_sequence(sequence:ndarray, n_step_in:int, n_step_out:int)->(ndarray, ndarray):

    x=[]
    y=[]

    for i in range(len(sequence)):

        if(i+n_steps_in+n_steps_out > len(sequence)):
            break

        x.append(sequence[i:i+n_steps_in, 0:2])
        y.append(sequence[i+n_steps_in:i+n_steps_in+n_steps_out, -1])

    return (np.asarray(x), np.asarray(y))






n_steps_in  = 3
n_steps_out = 2

(x, y) = split_sequence(dataset, n_steps_in, n_steps_out)

print(x.shape)
n_feature = x.shape[2]

 

3. 학습

from keras import Sequential
from keras.layers import Dense, Flatten
from keras.layers.convolutional import Conv1D, MaxPooling1D


model = Sequential()
model.add(Conv1D(64, 2, activation='relu', input_shape=(n_steps_in, n_feature)))
model.add(MaxPooling1D())
model.add(Flatten())
model.add(Dense(50, activation='relu'))
model.add(Dense(n_steps_out))
model.compile(optimizer='adam', loss='mse')

model.fit(x, y, batch_size=10, epochs=3000, verbose=1)

 

4. 표시

x_input = np.asarray([[70, 75],[80,85],[90,95]])    # expect [185, 205].
x_input = x_input.reshape((1, n_steps_in, n_feature))

yhat = model.predict(x_input, verbose=1)

print(yhat)  #[[208.41682 230.64803]]