Run U-Net in "MetNet mode": Predict one timestep at a time. Tell the model which timestep by using an input channel. Output a probability distribution. Only predict for the central crop.

openclimatefix / power_perceiver

Machine learning experiments using the Perceiver IO model to forecast the electricity system (starting with solar)

MIT License

7 stars 1 forks source link

Open JackKelly opened 2 years ago

JackKelly commented 2 years ago

Use this trick: https://twitter.com/shoyer/status/1521918728702550016
Re-ballance the bins by taking a cumulative histogram of image pixel values, and then dividing up the y axis into equal step sizes