What is Machine Learning?
Machine learning is the scientific study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions, relying on patterns and inference instead. It is seen as a subset of artificial intelligence.
· It is the application of Artificial Intelligence which provides or enables a system to learn the pattern automatically and find out the hidden information of the problem.
· There is a famous book “Machine learning in Action” by Peter Harrington where he has mentioned the basic knowledge about machine learning and its real-world application.
· ML is a technique which can solve the difficulties of all problems.
· ML is used to solve the complex pattern of any problem.
What is Artificial Intelligence?
AI is the other word of human intelligence processes by machines which is more efficient than humans and can solve what humans cannot do it.
· It draws pictures better than humans, write codes, getting more creative and innovative. M
· Both ML and AI go hand on hand. Both are related to each other.
From a set of observational data of features, class and labels, the machine learns the classification boundary between two or more different classes.
Machine learning pipeline:
The pipeline is an entire journey of machine learning.
1. How to Define a ML problem statement
2. How to gather and collect the data
3. Cleaning and processing of data
4. Extracting the features of data
5. How to learn to perform a task
6. Predict the output and implement it
Different types of ML Algorithm:
3 broad classifications-:
I. Supervised learning algorithm
II. Unsupervised learning algorithm
III. Reinforcement learning algorithm
1. Supervised learning: In this algorithm data is given by features as well as labels, where we can classify them as individually. example-: In gender data we can distinguish between male and female. Here we can do the prediction by seeing the historic set of data, which can provide more detail information of a data.
2. Unsupervised learning: In unsupervised algorithm data is only given by its features. Where there are no labels, no classifications. Sometimes its very much difficult to understand the mixture set of data where we cannot draw a clear straight line by seeing only one difference. We try to draw a line just to categorize them. So, we do clustering. It cannot give any specific output but it can only describe the hidden structure of a unlabeled data.
3. Reinforcement learning: It is a ML technique which is mainly concerned about with how software agents ought to take actions to maximize the profit, performance etc.