Implements a simple Linear SVM (soft margin formulation) trained in PyTorch by gradient descent on a 2D toy dataset with binary labels. The hyper-parameter search (C parameter) is done by cross-validation via scikit-learn. It includes a visualization of the decision function and associated support vectors throughout the epochs of training for various C values. Inspiration from:
- Python Engineer: https://youtu.be/UX0f9BNBcsY
- A Developer Diary: http://www.adeveloperdiary.com/data-science/machine-learning/support-vector-machines-for-beginners-linear-svm/
