Machine Learning
Ever thought how Google works, how recommendation of friends takes place, how self driving cars drives it self, how photo image recognition of your face gets added in any app such as Google Photos and Social Medias such as Facebook and LinkedIn. Yes, the answer is, Machine Learning we are in 5th generation of computers. Where computers play a major role in all fields and automatisation takes place every where, why so in computer science? I cant line tons and tons of code to solve a particular problem and once i need to add certain feature then the same process starts again. So how do I solve this? Remember we can’t assign the objects as this or that in world, for example we cant write code to identify apple form another fruit just in real life apple vary in colour and size are some exceptions. So the only way to solve this to write an algorithm that learns by its own without being explicitly programmed. Yes, we could make computer to learn things not just 1s and 0s. We can write a algorithm and use that to solve similar problems such as classification of apple, tumour and spam emails.
Classification of Machine Learning
1. Supervised Learning
2. Unsupervised Learning
there are other types too, Reinforced learning, and Recommender systems .
Supervised Learning, we could train the system with the correct set of data we have and use the system to predict or classify things. Example, List of prices of houses sold for last ten years and predict price of the house we could sell now.
- We analyse problem
- Pick or Create a model (Based on Problem)
- Train model based on the data set
- Implement the model and algorithm
In Unsupervised Learning, we don’t have any idea about the data set let the system find patterns and by itself come up with its own logic (For Simple Understanding) Let system do some mathematic stuff find things that differ and make a pattern and thats all about this. Example, group pixel in image.
Author : Dhyakesh Soundararajan
