1D CNN (Convolutional neural network)

[SverreNystad]

1D Convolutional Neural Networks (1D CNNs) can be highly effective for time series forecasting due to their ability to detect local temporal patterns and their capacity to handle sequential data.

1D CNNs can also be combined with other models like Recurrent Neural Networks (RNNs) or Long Short-Term Memory (LSTM) networks to leverage the advantages of both architectures. For example, a CNN can be used to extract local features, which can then be fed into an LSTM to capture long-term dependencies in the sequence.

1D CNNs offer several advantages for time series forecasting, including the ability to capture local temporal patterns, robustness to noise, and scalability to long sequences. Their efficiency, parallelizability, and ability to learn hierarchical features make them a popular choice for various time series applications.

[SverreNystad]

Code snippet for 1D CNN using KERAS

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv1D, MaxPooling1D, Flatten, Dense

model = Sequential()
model.add(Conv1D(filters=64, kernel_size=3, activation='relu', input_shape=(sequence_length, n_features)))
model.add(MaxPooling1D(pool_size=2))
model.add(Flatten())
model.add(Dense(50, activation='relu'))
model.add(Dense(1))
model.compile(optimizer='adam', loss='mse')

[SverreNystad]

Gave worse results