CNN-based-non-linear-regression

use 1d CNN model to train the non-linear function : y = x^2 + 6x + 9

• input x.shape:[1000, 1, 300]
• output y.shape:[1000, 1, 300]
• CNN model:
  • conv1d:(k_size=(300), stride=1, padding=148)
  • ReLU
  • conv1d:(k_size=(300), stride=1, padding=149)
  • ReLU