Generating synthetic Tensile Test daa

TensorFlow General Discussion
1

Hello all,

I am so new to TensorFlow and machine learning due to my background. Recently, I have started to look for TensorFlow to generate synthetic data for my project. I have done a tensile test on a newly developed alloy and got tensile stress-strain data from this test. Now, I am wondering if it is possible to use TensorFlow to generate 3 more tensile test data from this raw data set. I have read a couple of papers but I did not get any information. Therefore, I am wondering if anyone can help me with this. This will be a very big help for me.

Thank you

3

Hi @Huseyin_Hizli,

Apologize for the delay in response.
Yes, it is possible to generate a new dataset in tensorflow from an existing dataset.I’d suggest to go with generative models like Generative adversarial networks(GANs) which can generate the dataset that resemble the raw/existing dataset.Here’s a sample implementation gist of GAN for reference and check this documentation about GANs for more information.

Thank You.