Day 6 of Machine Learning||Supervised ML Algorithms

Neha Gupta - Jun 10 - - Dev Community

Hey reader😀Hope you are doing well🙂
In the last blog we have seen that how EDA is performed on a dataset. In this post we are going to discuss about Supervised Machine Learning and the algorithms that we can use to build our model.
So let's get started 🔥

What is Supervised Machine Learning?

Supervised machine learning is a type of machine learning where the algorithm learns from a labeled dataset.
Here the word labeled tells us that the dataset contains input as well as corresponding output values.
To understand it better let's take a dataset of Housing Price-:
Image description
We can see that here we have area, bedrooms, balcony, age and price as our columns. In this dataset the first four columns are determinant of price of a house. So we can say that price is dependent variable and rest are our independent variable. This is a labeled dataset.
We train our machine on the dataset so that it can make predictions for new, similar data.
Image description
Note that the above dataset contains numerical values only as output but we can have categorical data too. So we need to handle this very carefully to get accurate model.

Supervised Machine Learning Problems

Based on the type of output column in our dataset, we have two types of supervised ML problem-:

  • Regression Problem

  • Classification Problem

Regression Problem

Regression is generally used for datasets where the output is continuous. The above example is a type of regression problem. The algorithms used include Linear Regression ,Locally Weighted Regression,Lasso Regression ,Ridge Regression ,Decision Trees, Neural Networks, Random Forest etc.

Classification Problem

Classification is generally used for datasets where the output is discreet. For example based on some data a candidate can either lose the election or win the election. The algorithms used include Logistic Regression, Support Vector Machine, Decision Trees, Random Forest, Naive Bayes, Neural Networks etc.

So now we are aware of the type of problems that we can encounter in Supervised Learning. In the next blog we are going to study about our very first algorithm i.e. Linear Regression.
Hope you have understood and liked this blog. Don't forget to like it and follow me.
Thankyou ❤️

. . . . . . . . . . . . . . . . . . . . . . .
Terabox Video Player