Types of Machine Learning

We are living in a century in which we have done some really ground breaking inventions. A century where computing moved from large mainframes to PCs to cloud machine and now to serverless computing. All this is really some amazing State of the art but it is just the beginning of what is to come next.

The biggest problem for AI researchers was the limitation of computing powers but thanks to the great minds now we have machines which are really really fast and also we have cloud which gives us many cool feature and computing power.

So now moving to machine learning, all the algorithms can be broadly categorised in 3 categories:-

1. Supervised learning:- It consists of target variables which are to be protected from input variables. Algorithms uses of function which maps this input variable to the output. This process is called training and it continues until undefined error rate or accuracy is achieved. Example of supervised learning are regression, decision trees, KNN.
These algorithms require input data to be preprocessed for better training. Usually this is the most time consuming process in machine learning.

2. Unsupervised learning:- In this algorithm we don't have any target variable. This is mostly used for clustering like dividing population into segments, detecting anamolies in network. Example of Unsupervised learning are K means algorithm, Apriori algorithm.

3. Reinforcement learning:- This algorithm learns on the basis of the feedback from the environment it is running or training. It learns from past experience and tries to capture best knowledge to make accurate decisions. Example of reinforcement learning are Markov Decision process, Monte Carlo.

We will study all 3 types of algorithms in future.

If you have any doubt then please post it in comments and I will try to answer is best I can.

Cheers and enjoy Machine Learning!!!