### Introduction to Support Vector System

As we know, Support Vector System is a supervise learning way and a machine learning algorithm. Moreover, we can use it for classification and regression problems. Certainly, it is an acronym for SVM.

Firstly, in SVM, we plot the data item in a n-dimensional space with value of each feature being a particular coordinate. Secondly, we perform classification by finding hyper plane which differentiates the classes as well. Thirdly, it creates coordinates of each individual observation.

The main objective of Support Vector System is to create best line boundary which can segregate n dimensional space into various classes. So we can put new data points easily. Hyper plane is the best decision boundary.

Moreover, it chooses vectors or extreme points which help in creating hyper plane. These vectors are the support vectors.

In addition, there are two types of SVM:

- Linear SVM
- Non linear SVM

## Table of Contents

- How does Support Vector System work?
- Pros of using SVM
- Cons of using SVM
- Conclusion

### How does Support Vector System work?

- Firstly, we need to identify the right hyper plane. For instance, there are three hyper planes. The rule is that select that plane which segregates the two classes much better.
- Secondly, another way is to maximize the distance between nearest data point and the plane.
- It is easy to have a linear hyper plane in SVM. Moreover, it has a technique called the kernel trick. It takes low dimensional input and converts it to higher dimension space.

To implement this algorithm in Python, we can use scikit-learn. Moreover, if you want to implement SVM in R, you can use e1071 package. There are functions for Naive Bayes Classifier in this package.

### Pros of using SVM

- This algorithm is highly efficient for high dimensional spaces.
- It works well with a clear margin of separation.
- It works nicely when number of dimensions becomes greater than number of samples.
- Support vectors are memory efficient.

### Cons of using SVM

- Not an accurate algorithm for larger data set.
- It does not work well with noisy data and when target classes overlaps.
- There are no probability estimates.

## Conclusion

In conclusion, we have learnt that Support Vector System is a supervise learning way and a machine learning algorithm. Moreover, we can use it for classification and regression problems. Certainly, it is an acronym for SVM.

Firstly, we need to identify the right hyper plane. Second way is to maximize the distance between nearest data point and the plane. To implement this algorithm in Python, we can use scikit-learn and in R, we use e1071 package.

This algorithm is highly efficient for high dimensional spaces. Also, support vectors are memory efficient. We have also seen at some of the major disadvantages of using SVM.