Intro to Machine Learning.......

**What is machine learning?** It is the study of computer algorithm that improve automatically through experience and by the use of data. Basically we train machine on the data set to perform a operation. For simple tasks, we can program the steps and there is no learning required. But for advance tasks it's hard to develop such program**.** **Types of Machine Learning** **Supervised Learning** :- learning a predicted function that maps input with output using labelled data. It is further categorized into **Classification** and **Regression.** In both cases we are giving examples of desired output i.e we are showing it beforehand what it had to predict.

**Unsupervised Learning** :- this type of ML looks for patterns in dataset with no pre-existing labels. Example:- In google news similar type of news are grouped together. It is called **Clustering**, in it we cluster similar type of data. Another type of unsupervised learning is **Association rule**. It is used in recommendation system. It find pattern in data and find association rules.

**Semi-supervised Learning** :- when we combine the aspects of labeled data (supervised) and un-labeled (unsupervised) data we get semi-supervised learning. in it we combine small amount of labeled data. this is used in fraud detection technique. **Reinforcement Learning** :- it has two components agent and environment. It is used in online Chess. the agent is the system that plays against us and environment is the other component that affects its next moves. this sytem learns through feedback, for good move it get +ve points and vice versa. **Steps in ML project** :- Data Collection and Pre-processing Training Deployment of model

**to be continued........................ next time in ML:- Hand written digit recognition**