K Means cluster
This a unsupervised machine learning algorithm. It form data into K clusters. Cluster are formed by choosing centroid. At first random centroid is picked. Iterating the process until centroid changes no more.
In this algorithm flowchart of decision is formed on basis of that prediction is done. Every feature chosen as such it decreases entropy.
How different our data is or how variation present in our data is called entropy.
SVM – Support Vector Machine
When number of features are more then SVM is a good choice. It creates hyperplanes which separate various cluster of data. It is supervised learning algorithm. It works on kernel. Different kernels can be used as convenience.
It is simple combine two or more models results together. For example – Random Forest – Decision trees are bagged here, means result from various Decision tree is averaged.