Apakah ada metode untuk menghitung interval prediksi (distribusi probabilitas) di sekitar perkiraan deret waktu dari jaringan saraf LSTM (atau berulang lainnya)?
Katakanlah, misalnya, saya memprediksi 10 sampel ke masa depan (t + 1 hingga t + 10), berdasarkan pada 10 sampel terakhir yang diamati (t-9 hingga t), saya berharap prediksi pada t + 1 menjadi lebih akurat daripada prediksi di t + 10. Biasanya, orang mungkin menggambar bar kesalahan di sekitar prediksi untuk menunjukkan interval. Dengan model ARIMA (dengan asumsi kesalahan berdistribusi normal), saya dapat menghitung interval prediksi (misalnya 95%) di sekitar setiap nilai prediksi. Bisakah saya menghitung yang sama, (atau sesuatu yang berhubungan dengan interval prediksi) dari model LSTM?
Saya telah bekerja dengan LSTM di Keras / Python, mengikuti banyak contoh dari machinelearningmastery.com , dari mana kode contoh saya (di bawah) didasarkan pada. Saya sedang mempertimbangkan membingkai ulang masalah sebagai klasifikasi ke dalam tempat sampah terpisah, karena hal itu menghasilkan kepercayaan per kelas, tetapi itu tampaknya solusi yang buruk.
Ada beberapa topik serupa (seperti di bawah), tetapi sepertinya tidak ada yang secara langsung mengatasi masalah interval prediksi dari LSTM (atau memang lainnya) jaringan saraf:
/stats/25055/how-to-calculate-the-confidence-interval-for-time-series-prediction
Prediksi seri waktu menggunakan ARIMA vs LSTM
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import LSTM
from math import sin
from matplotlib import pyplot
import numpy as np
# Build an LSTM network and train
def fit_lstm(X, y, batch_size, nb_epoch, neurons):
X = X.reshape(X.shape[0], 1, X.shape[1]) # add in another dimension to the X data
y = y.reshape(y.shape[0], y.shape[1]) # but don't add it to the y, as Dense has to be 1d?
model = Sequential()
model.add(LSTM(neurons, batch_input_shape=(batch_size, X.shape[1], X.shape[2]), stateful=True))
model.add(Dense(y.shape[1]))
model.compile(loss='mean_squared_error', optimizer='adam')
for i in range(nb_epoch):
model.fit(X, y, epochs=1, batch_size=batch_size, verbose=1, shuffle=False)
model.reset_states()
return model
# Configuration
n = 5000 # total size of dataset
SLIDING_WINDOW_LENGTH = 30
SLIDING_WINDOW_STEP_SIZE = 1
batch_size = 10
test_size = 0.1 # fraction of dataset to hold back for testing
nb_epochs = 100 # for training
neurons = 8 # LSTM layer complexity
# create dataset
#raw_values = [sin(i/2) for i in range(n)] # simple sine wave
raw_values = [sin(i/2)+sin(i/6)+sin(i/36)+np.random.uniform(-1,1) for i in range(n)] # double sine with noise
#raw_values = [(i%4) for i in range(n)] # saw tooth
all_data = np.array(raw_values).reshape(-1,1) # make into array, add anothe dimension for sci-kit compatibility
# data is segmented using a sliding window mechanism
all_data_windowed = [np.transpose(all_data[idx:idx+SLIDING_WINDOW_LENGTH]) for idx in np.arange(0,len(all_data)-SLIDING_WINDOW_LENGTH, SLIDING_WINDOW_STEP_SIZE)]
all_data_windowed = np.concatenate(all_data_windowed, axis=0).astype(np.float32)
# split data into train and test-sets
# round datasets down to a multiple of the batch size
test_length = int(round((len(all_data_windowed) * test_size) / batch_size) * batch_size)
train, test = all_data_windowed[:-test_length,:], all_data_windowed[-test_length:,:]
train_length = int(np.floor(train.shape[0] / batch_size)*batch_size)
train = train[:train_length,...]
half_size = int(SLIDING_WINDOW_LENGTH/2) # split the examples half-half, to forecast the second half
X_train, y_train = train[:,:half_size], train[:,half_size:]
X_test, y_test = test[:,:half_size], test[:,half_size:]
# fit the model
lstm_model = fit_lstm(X_train, y_train, batch_size=batch_size, nb_epoch=nb_epochs, neurons=neurons)
# forecast the entire training dataset to build up state for forecasting
X_train_reshaped = X_train.reshape(X_train.shape[0], 1, X_train.shape[1])
lstm_model.predict(X_train_reshaped, batch_size=batch_size)
# predict from test dataset
X_test_reshaped = X_test.reshape(X_test.shape[0], 1, X_test.shape[1])
yhat = lstm_model.predict(X_test_reshaped, batch_size=batch_size)
#%% Plot prediction vs actual
x_axis_input = range(half_size)
x_axis_output = [x_axis_input[-1]] + list(half_size+np.array(range(half_size)))
fig = pyplot.figure()
ax = fig.add_subplot(111)
line1, = ax.plot(x_axis_input,np.zeros_like(x_axis_input), 'r-')
line2, = ax.plot(x_axis_output,np.zeros_like(x_axis_output), 'o-')
line3, = ax.plot(x_axis_output,np.zeros_like(x_axis_output), 'g-')
ax.set_xlim(np.min(x_axis_input),np.max(x_axis_output))
ax.set_ylim(-4,4)
pyplot.legend(('Input','Actual','Predicted'),loc='upper left')
pyplot.show()
# update plot in a loop
for idx in range(y_test.shape[0]):
sample_input = X_test[idx]
sample_truth = [sample_input[-1]] + list(y_test[idx]) # join lists
sample_predicted = [sample_input[-1]] + list(yhat[idx])
line1.set_ydata(sample_input)
line2.set_ydata(sample_truth)
line3.set_ydata(sample_predicted)
fig.canvas.draw()
fig.canvas.flush_events()
pyplot.pause(.25)