K-Means Clustering Algorithm
k-means is one of the simplest unsupervised learning algorithms that solve the well known clustering problem. It is a non-deterministic and iterative method. Its algorithm works over a specified data set using a predefined set of clusters, k. The result of a k-means algorithm is the k clusters of input data partitioned from the clusters. For eg. let’s have a k-Means clustering application for Wikipedia search results. If we use the search term Lincoln over Wikipedia, we will get results containing this word, and they usually give us references to the past U.S. president, a city in the United States, or a car company. We can use this algorithm to put these results and pages into groups which talk about the same concepts. Thus, the algorithm will group the results and pages into clusters. A k-means clustering algorithm is commonly used by search engines to identify similarities of search results. This algorithm provides lesser time for the searcher when looking for a more precise result using the search engine.
A Marketing Example
K-means clustering stands out to be an effective tool when developing marketing and buying personas out of clusters of information concerning buyers. Cluster analysis proves to be useful for market segmentation. This analysis means having to divide a market’s potential customers over to different subsets wherein customers that belong to the same group are the same concerning a specified characteristic and vice versa. This arrangement will provide a convenient and more accurate marketing mix depending on the target customer specified.