Kernelization is a technique used to design efficient algorithms by preprocessing inputs and replacing them with smaller inputs called "kernels". Reduction rules are applied to the instance to cut away parts that are easy to handle, and if successful, results in a fixed-parameter tractable algorithm with polynomial time kernelization step and non-polynomial but bounded time to solve the kernel.
Carnegie Mellon University
Spring 2018
A comprehensive exploration of machine learning theories and practical algorithms. Covers a broad spectrum of topics like decision tree learning, neural networks, statistical learning, and reinforcement learning. Encourages hands-on learning via programming assignments.
No concepts data
+ 55 more concepts