A Gradient-Descent-Based k-NN Algorithm
Wenqi Guo · 2022
In this paper, we reviewed a few works that solved the speed issue of the k-NN classification algorithm. This makes it possible to solve real problems using the classificational k-NN algorithm. We then use gradient descent to achieve the equivalent goal of feature scaling, increasing the accuracy of the k-NN classification algorithm, including on nominal data. I was just a part-time college student when I wrote this paper, things might be wrong. Don't assume its credibility is high.