CNN 帮助（非常新）

Question

我正在制作一个深度学习模型来预测一个值。声明如下：我们有一个
时间 vs A vs B vs C - xls 文件（多个，存在于 Data 命名文件夹中）
我想使用 A、B、C 的先前值来预测 A。这意味着当它是第一个预测时，我们将得到 N/A，因为我们没有旧的 A 给用户。 目前我有这个：（我真的是新手，到处都在网上搜索）：

def load_data(folder_path):
    data_frames = []
    for file_name in os.listdir(folder_path):
        if file_name.endswith('.xls'):
            file_path = os.path.join(folder_path, file_name)
            df = pd.read_excel(file_path)  # Assuming the data is in Excel format
            data_frames.append(df)
    return pd.concat(data_frames, ignore_index=True)

def preprocess_data(data):
    X = data[['S', 'A', 'T']]
    y = data['F']
    scaler = StandardScaler()
    X_scaled = scaler.fit_transform(X)
    X_reshaped = X_scaled.reshape(X_scaled.shape[0], X_scaled.shape[1], 1)  # Reshape for CNN
    return X_reshaped, y, scaler


def build_cnn_model(input_shape):
    model = Sequential()
    model.add(Conv1D(filters=1, kernel_size=10, activation='relu', input_shape=(1, 1,3),))
    model.add(MaxPooling1D(pool_size=2))
    model.add(Conv1D(filters=64, kernel_size=3, activation='relu'))
    model.add(MaxPooling1D(pool_size=2))
    model.add(Flatten())
    model.add(Dense(64, activation='relu'))
    model.add(Dense(1)) 
    return model



def train_and_predict_next_frequency(folder_path, current_scheduled, current_actual):
    data = load_data(folder_path)

    X, y, scaler = preprocess_data(data)

    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

    input_shape = (X_train.shape[1], 1)  
    X_train_reshaped = X_train.reshape(X_train.shape[0], X_train.shape[1], 1)
    X_test_reshaped = X_test.reshape(X_test.shape[0], X_test.shape[1], 1)

    print(input_shape)
    model = build_cnn_model(input_shape)

    model.compile(optimizer='adam', loss='mean_squared_error')

    model.fit(X_train_reshaped, y_train, epochs=50, batch_size=32, verbose=1)

    mse = model.evaluate(X_test_reshaped, y_test, verbose=0)
    print("Mean Squared Error:", mse)

    scaled_input = scaler.transform([[current_scheduled, current_actual]])
    input_reshaped = scaled_input.reshape(1, 2, 1)  # Reshape for CNN
    next_frequency_scaled = model.predict(input_reshaped)
    next_frequency = scaler.inverse_transform(next_frequency_scaled.reshape(-1, 1))
    return next_frequency[0][0]



folder_path = "Data"  # Folder containing the Excel files
current_scheduled = 10
current_actual = 8
predicted_frequency = train_and_predict_next_frequency(folder_path, current_scheduled, current_actual)
print("Predicted Next Frequency:", predicted_frequency)

我在这里收到此错误：ValueError：

One of the dimensions in the output is <= 0 due to downsampling in conv1d_10. Consider increasing the input size. Received input shape [None, 1, 1, 3] which would produce output shape with a zero or negative value in a dimension.

请帮忙。预先感谢，因为我知道这是一篇 LONGGG 帖子:(

使用

A、B、C
的先前值来预测 A。 （上一个是什么？，我们有时间来告诉）。 我尝试更改图层功能，但没有成功。我想实现上述目标

Answer 1

padding='same'

添加到

Conv1D

应该可以解决该错误。model.add(Conv1D(filters=1, kernel_size=10, activation='relu', input_shape=(1, 1,3),padding='same'))