ConvGeNCode/documentation/paper/Benchmark/Algorithms/convGAN.md

convGAN

Is our approach for systetic point generation for imbalanced datasets.

In our benchmark we use the following settings:

• proximary in majority to minority / full majority class
• Neighborhood size: 5 / all points in the majority data set

(GIT link: [https://github.com/kristian10007/convGAN])