search

lululxvi / deep-learning-for-indentation

Extraction of mechanical properties of materials through deep learning from instrumented indentation
Apache License 2.0
55 stars 27 forks source link

High MAPE with DeepXDE 1.1.2 and TensorFlow 2.7.0 – Seeking Assistance #16

Open jingyu124 opened 8 months ago

jingyu124 commented 8 months ago

Hello Professor, I am trying to present the results of Figure S1 A from your paper, specifically focusing on the values of E*.

I am using DeepXDE version 1.1.2 with TensorFlow 2.7.0 as the backend. The data is generated by executing the gen-forward code in model.py.

Is the MAPE value on the Y-axis of Figure S1 directly calculated in the code, or should it be separately computed from the result files? The average MAPE values are excessively high. Can you suggest any potential issues causing this? Thank you.

image image

lululxvi commented 8 months ago

Yes, it should be correct. I am not sure if it is the problem of TF version, or something else.

You can look at simple examples at https://deepxde.readthedocs.io/en/latest/demos/function.html#multi-fidelity-learning