Question

Define the k-nearest neighbor and the neural network classification technique (definition only).

Answer 1

Ques- A k-nearest neighbor is an information arrangement calculation that endeavors to figure out what bunch an information point is in by taking a gander at the information focuses around it. A calculation, seeing one point on a framework, attempting to decide whether a point is in bunch An or B, takes a gander at the conditions of the focuses that are close to it. The range is self-assertively decided, however the fact is to take an example of the information. On the off chance that most of the focuses are in bunch An, at that point all things considered, the information point being referred to will be An as opposed to B, and the other way around. The k-closest neighbor is a case of a "languid student" calculation since it doesn't create a model of the informational collection previously. The main computations it makes are the point at which it is approached to survey the information point's neighbors. This makes k-nn extremely simple to execute for information mining.

A neural network classification comprises of units (neurons), masterminded in layers, which convert an information vector into some yield. Every unit takes an information, applies a (regularly nonlinear) capacity to it and afterward gives the yield to the following layer.