The notebook, SVMKernels.ipynb, displays my implementation of a kernel-based Support Vector Machine. The SVM class is implemented in the notebook while supporting classes and code was implemented in source.py. The source code was extracted from the notebook inorder to maintain readability and keep the notebook focused on the SVM itself rather than other required components (i.e. One vs Rest classifier, parallel progress bar, etc). However, I find both files quite interesting for various reasons.
Please take a look and feel free to email me with any questions or comments! I'd love to hear some feedback.
The notebook implements parallel computing techniques that utilize the forking mechanism embedded into Linux based kernels and as such won't work on Windows. Because of this, it is hosted and can be found at Azure Notebooks. For full functionality, the kernel must be running with Python version 3.6. Enjoy!