=================================================================================

K-Nearest Neighbor (KNN) is one of the fundamental algorithms in machine learning. Machine learning models use a set of input values to predict output values. KNN is one of the simplest forms of machine learning algorithms mostly used for classification. It classifies the data point on how its neighbor is classified. KNN can classify the new data points based on the similarity measure of the earlier stored data points. K in KNN represents the number of the nearest neighbors we used to classify new data points. Choosing the right value of K is called parameter tuning and it’s necessary for good results. However, odd value of K is always selected to avoid confusion between 2 classes.

K-Nearest Neighbours (KNN Algorithm) (see page4707) can be used for identification of global defects and local defects in defect denoising. This can be done by, for instance, assuming that defects were distributed according to two superimposed Poisson point processes:
         i) Global defects as a Poisson point process with intensity λ₁,
         ii) Local defects as a Poisson point process with intensity λ₂ (> λ₁), but restricted to some regions of the wafer and overlaid on the global defects. [1] The k^th NN distance D_k of a randomly chosen point defect is its distance to its k^th NN defect.

Note if multiple local defect clusters have different defect densities, this NN clutter removal denoising approach may not perform well, and we found that the spatial filter [2, and see page4256] may perform better.

Table 4255a. Applications and related concepts of nearest-neighbor (NN).

============================================

[1] Tao Yuan, Way Kuo, Suk Joo Bae, Detection of Spatial Defect Patterns Generated in Semiconductor Fabrication Processes, IEEE Transactions on Semiconductor Manufacturing, 24(3), DOI: 10.1109/TSM.2011.2154870, 2011.
[2] C. H. Wang, W. Kuo, and H. Bensmial, “Detection and classification of defects patterns on semiconductor wafers,” IIE Trans., vol. 39, no. 12, pp. 1059–1069, 2006.