There are many different ways we can classify types of machine learning. In this post we are going to classify based on the amount and type of supervision these machine learning systems get during training. This classification is called Supervised/Unsupervised Learning. Following are the four major paradigms:
Supervised Learning: Training data includes the desired solutions called labels. Important algorithms that come under supervised learning are: Linear Regression, Logistic Regression, Decision Trees, Random Forests, k-Nearest Neighbors, some Neural Networks, etc.
Unsupervised Learning: Training data is unlabeled. Important algorithms that come under unsupervised learning are: Principal Component Analysis(PCA), k-Means Clustering, Hierarchical Cluster Analysis, Apriori, etc.
Semisupervised Learning: Algorithms that can deal with partially labeled data, little bit of labeled data and lot of unlabeled data. Example: Restricted Boltzmann machines(RBMs) are trained sequentially in unsupervised manner and then fine-tuned using supervised learning.
Reinforcement Learning: An agent learns by itself what is the best strategy called policy to get the most reward over time.
We will explore each of the paradigms in the later blog posts. I have already used some of the algorithms to train different models. You can find it here.