In layman’s language, Machine Learning as the name suggests, involves “making the machine learn”.
Learn what?
Learn the decisions to be taken on a particular case, just like humans do. Hence, Machine Learning is a branch of Artificial Intelligence.
Learn How?
How do humans learn something?
By practice, i.e by applying your knowledge many times on different cases.
Imagine you’ve joined Tennis coaching classes. Your coach tells the assistant coach shoot balls at you in different directions and with different speeds. He spins the ball sometimes. He might also shoot the ball at you in ways which is rare. For every ball thrown at you, the coach tells you to hit the ball in a particular way. Maybe for a straight shot with less speed he will tell you to hit it towards the opponents net, while away from the nets for some other shot. In this way you’ll keep practicing on various shots and learn tennis one day. After you’ve learnt tennis, you’ll be able to hit the ball without the help of your coach.
Now,
You- Learning Model
Assistant Coach- Training Data
Coach- Decisions on Training DataAfter Learning,
You- Trained Model able to take decisions on Test Data
In machine learning, we give the algorithmic model some training data which already has the decisions taken. The model learns from the data using the algorithm. Next time when you give the model an unknown case, it will make the decision.
There are various ways through which you can learn machine learning, but I think taking this Stanford MOOC by Andrew Ng will be the most helpful way. Apart from being the most popular MOOC, it covers all the topics in Machine Learning with informative videos. It also has multiple quizzes and programming assignments to test what you’ve learnt. Machine Learning – Coursera Machine Learning
You’re gonna love Machine Learning 🙂
– Aditya Awalkar
shuntzblog 