我正在嘗試為多步預(yù)測(cè)創(chuàng)建 LSTM 模型。現(xiàn)在我正在測(cè)試模型網(wǎng)絡(luò)設(shè)置，但發(fā)現(xiàn)它在設(shè)置上存在尺寸問(wèn)題。這是我的測(cè)試數(shù)據(jù)集：length = 100df = pd.DataFrame()df['x1'] = [i/float(length) for i in range(length)]df['x2'] = [i**2 for i in range(length)]df['y'] = df['x1'] + df['x2'] x_value = df.drop(columns = 'y').valuesy_value = df['y'].values.reshape(-1,1)這是我的 t 窗口數(shù)據(jù)構(gòu)建函數(shù)：def build_data(x_value, y_value ,n_input, n_output): X, Y = list(), list() in_start = 0 data_len = len(x_value) # step over the entire history one time step at a time for _ in range(data_len): # define the end of the input sequence in_end = in_start + n_input out_end = in_end + n_output if out_end <= data_len: x_input = x_value[in_start:in_end] # e.g. t0-t3 X.append(x_input) y_output = y_value[in_end:out_end] # e.g. t4-t5 Y.append(y_output) # move along one time step in_start += 1 return np.array(X), np.array(Y) X, Y = build_data(x_value, y_value, 1, 2)X 和 Y 的形狀X.shape### (98, 1, 2)Y.shape### (98, 2, 1)對(duì)于模型部分，verbose, epochs, batch_size = 1, 20, 16n_neurons = 100n_inputs, n_features = X.shape[1], X.shape[2]n_outputs = Y.shape[1]model = Sequential()model.add(LSTM(n_neurons, input_shape = (n_inputs, n_features), return_sequences=True))model.add(TimeDistributed(Dense(1)))model.compile(loss='mean_squared_error', optimizer='adam')model.fit(X, Y, epochs=epochs, batch_size=batch_size, verbose=verbose)它發(fā)生了錯(cuò)誤： ValueError: Error when checking target: expected time_distributed_41 to have shape (1, 1) but got array with shape (2, 1)如果使用那X, Y = build_data(x_value, y_value, 2, 2) i.e. input window == output window將是有效的。但我認(rèn)為它不應(yīng)該包含這個(gè)約束。我怎樣才能克服這個(gè)問(wèn)題？即設(shè)置為input window != output window或我應(yīng)該設(shè)置的任何圖層或設(shè)置？
1 回答
蝴蝶不菲
TA貢獻(xiàn)1810條經(jīng)驗(yàn) 獲得超4個(gè)贊
您在處理時(shí)間維度時(shí)遇到形狀不匹配...當(dāng)您嘗試預(yù)測(cè)時(shí)間維度為 2 的內(nèi)容時(shí)，時(shí)間輸入 dim 為 1。因此您的網(wǎng)絡(luò)中需要一些能夠從 1 增加到 2 時(shí)間的東西方面。我使用了Upsampling1D圖層，下面是一個(gè)完整的例子
# create fake data
X = np.random.uniform(0,1, (98,1,2))
Y = np.random.uniform(0,1, (98,2,1))
verbose, epochs, batch_size = 1, 20, 16
n_neurons = 100
n_inputs, n_features = X.shape[1], X.shape[2]
n_outputs = Y.shape[1]
model = Sequential()
model.add(LSTM(n_neurons, input_shape = (n_inputs, n_features), return_sequences=True))
model.add(UpSampling1D(n_outputs))
model.add(TimeDistributed(Dense(1)))
model.compile(loss='mean_squared_error', optimizer='adam')
model.fit(X, Y, epochs=epochs, batch_size=batch_size, verbose=verbose)
輸入時(shí)間暗淡 > 輸出時(shí)間暗淡，您可以使用 Lambda 或 Pooling 操作（如果維度匹配）。下面是一個(gè)使用 Lambda 的例子
X = np.random.uniform(0,1, (98,3,2))
Y = np.random.uniform(0,1, (98,2,1))
verbose, epochs, batch_size = 1, 20, 16
n_neurons = 100
n_inputs, n_features = X.shape[1], X.shape[2]
n_outputs = Y.shape[1]
model = Sequential()
model.add(LSTM(n_neurons, input_shape = (n_inputs, n_features), return_sequences=True))
model.add(Lambda(lambda x: x[:,-n_outputs:,:]))
model.add(TimeDistributed(Dense(1)))
model.compile(loss='mean_squared_error', optimizer='adam')
model.fit(X, Y, epochs=epochs, batch_size=batch_size, verbose=verbose)
添加回答
舉報(bào)
0/150
提交
取消