Part 1 - SVM with R | Supervised Learning | Kernlab package | ksvm | ML | Analytics with R

Data Set: https://archive.ics.uci.edu/ml/datase...

Support Vector Machine (SVM) is a powerful machine learning algorithm that can be used for both classification and regression tasks. The SVM algorithm tries to find the optimal hyperplane in an N-dimensional space that can separate the data points in different classes in the feature space. The hyperplane tries that the margin between the closest points of different classes should be as maximum as possible.

The images used in the video are picked from Google to explain core concepts and I do not own them.

Video summary [00:00:01]- [00:15:07]:

This video is a tutorial on how to use support vector machines (SVMs) in R. It covers the basics of SVMs, the kernel trick, the cost parameter, and how to apply SVMs to a credit card approval data set using the kernlab package. It also compares the performance of linear and non-linear kernels on the data set.

**Highlights**:
[00:00:01] *Introduction to SVMs*
Supervised machine learning algorithm
Takes input data and responses to form a model
Separates data into two categories using a hyperplane
[00:00:33] *Linear kernel*
Used when data is linearly separable
Maximizes the margin between the data points
Avoids overfitting the training data
[00:00:54] *Kernel trick*
Used when data is not linearly separable
Projects data into higher dimensions
Separates data using a non-linear kernel
[00:02:00] *RStudio demonstration*
Uses pacman and kernlab packages
Reads credit card approval data set
Splits data into training and testing sets
[00:03:00] *Cost parameter*
Controls the width of the margin
Varies from 10^-5 to 10^5
Affects the accuracy of the predictions
[00:04:00] *Linear SVM model*
Uses ksvm function with linear kernel
Gets the equation of the hyperplane
Predicts the test data using the model
[00:06:00] *Non-linear SVM model*
Uses ksvm function with Gaussian kernel
Gets the equation of the hyperplane
Predicts the test data using the model
[00:08:00] *Comparison of models*
Uses a table to show the accuracy of each model
Finds the optimal cost parameter for each model
Concludes that linear kernel performs better

Related Videos:

Analytics with R playlist:    • Analytics with R  
Install R and RStudio | Analytics | ML | Windows :    • Install R and RStudio | Analytics | M...  
SVM with R | Supervised Learning | Kernlab package | ksvm | ML | Analytics with R :    • Part 1 - SVM with R | Supervised Lear...  
knn with R | Supervised Learning | kknn package | ML | Analytics with R:    • Part 2 - knn with R | Supervised Lear...  
kknn & CV with R | Cross Validation | Supervised Learning | ML | Analytics with R :    • Part 3 - kknn & CV with R | Cross Val...  
KMeans with R | Unsupervised ML | ggplot | confusion matrix | Analytics with R :    • Part 4 - KMeans with R | Unsupervised...  
Detecting Outliers in dataset | Grubbs test | Box Plot | ML | Analytics with R :    • Part 5 - Detecting Outliers in datase...  
Trends in time series data | Holt-Winters method | rollmean | ML | Analytics with R :    • Part 6 - Trends in time series data |...  
Linear Regression | lm | glm | cross validation | Machine Learning (ML) | Analytics with R :    • Part 7- Linear Regression | lm | glm ...  
Principal Component Analysis (PCA) | prcomp | lm | R2 r squared | ML | Analytics with R :    • Part 8 - Principal Component Analysis...  
Logistic Regression | glm | binomial | ROC | AUC | Machine Learning (ML) | Analytics with R :    • Part 9 - Logistic Regression | glm | ...  
Regression Decision Trees | tree | pruning | Supervised Machine learning | Analytics with R :    • Part10 - Regression Decision Trees | ...  
Random forest | Classification | randomforest | Machine Learning (ML) | Analytics with R :    • Part 11 - Random forest | Classificat...  

#supportvectormachine #machinelearning #ml #r #supervisedlearning