Machine learning algorithm: “K-MEANS” clustering under Unsupervised learning

K-MEANS is an unsupervised type algorithm. It belongs to the Clustering type.

K-MEANS = clustering will partition the N datapoints or Observations, into K number of clusters[groups].

K-MEANS will create multiple number of Clusters using all input Observations.

If we want to predict Observation X is part of which Cluster, then send this new Observation X as input to K-Means model, which will return the Clustername as output.

Example: If we have 100 input observations of Customers profiles.

The K-MEANS cluster might create these Clusters:

  1. clustername “Rich” = 30 Observations [customers]
  2. clustername “Poor” = 70 Observations [customers]

I want to predict whether the new Customer observation dataset is a “Rich” or “Poor” type of Customer.

I will pass the new Customer observation data to my K-MEANS model and this will return “Rich” or “Poor” with accuracy rate %.

**This is a simple explanation of K-MEANS algorithm.