Try to fit the R^3 -> R^3 five dimensional function with a neural network.
You are free to choose the architecture and training method you want.

The samples you have access to have NOT been noised.


## Files

*   **3D_3D.csv** - the training set
*   **3D_3D-test_set.csv** - the test set
*   **3D_3D-submission_example.csv** - a sample submission file in the correct format