This program is used to forecast upcoming tool wear by LSTM

The speed is ultra-fast and final score is beyond satisfactory.

You can click [HERE] to know our idea, work and achievement related to this repo.

This slide is localized in Chinese.

Also, HERE is this project's homepage. You can find hints at our homepage.

1st,2nd and 3rd is corresponding to forecasting three timestep

MAE is lower than 1 and result is beyond satisfactory

In addition, we have also test how to make RNN chained for long-term tool wear forecasting. If you feel interest, you can follow the video in seq2seq_tool_wear

which named as

• cut_1.mp4
• cut_2.mp4
• cut_3.mp4

• cut_1 : https://youtu.be/arKw_D7j5pc
• cut_2 : https://youtu.be/n-Pyet1UJEg
• cut_3 : https://youtu.be/CANADaKSgac

Anyway, please feel free to ask me if you need further help

MIT LICENCE