file:$///$C:$/$Users$/14093/$OneDrive$/$credit$2.$pdf Use Rstudio and give $3$ sentences explaining each step of code. This assignment asks you to examine the Support Vector Machine for classification. Provide your answers to the questions in a Word document named Assign$4\_$LastName.doc along with your source code that should be saved as Assign$4\_$LastName, and click the title link to upload and submit them.

The attached dataset contains German credit data. To apply Support Vector Machines, the data requires preprocessing, such as data type transformation and normalization. For data type transformations, we mainly perform factoring of the categorical variables, where we transform the data type of the categorical features from numeric to factor.For example,

code line $5$: #data preprocessing; code line $6$: #data type transformation $-$ factoring; coding line $7$: to$.$Factor $<-$ function $($df$,$ variables$)\{$; coding line $8$: for $($variable in variables$)\{$; coding line $9$: df$[[$variable$]]<\_$as$.$factor$($df$[[$variable$]])$; coding line $10$: $\}$; coding line $11$: return $($df$)$; coding line $12$: $\}$

There are several numeric variables, which include credit.amount, age, and credit.duration.months. Please check their distributions using histogram. If they are skewed distributions, please normalize the data. One possible way is using z$-$normalization as follows:

coding line $14$: #Normalization$-$scaling; coding line $15$: scales.features $<\_$function$($df$,$ variables$)\{$; coding line $16$: for $($variable in variables$)\{$; coding line $17$: df$[[$variable$]]<\_$scale$($df$[[$variable$]],$ center$=$T$,$ scale$=$T$)$; coding line $18$: $\}$; coding line $19$: return$($df$)$; coding line $20$: $\}$

You can pass some variables to the above functions to transform and normalize the data, such as the following example:

code: #Normalize variable numeric.var $<-$c$("$your variables"$0$ yourData $<-$ scale.features$($yourData$,$ numeric.var$)$

You can apply a similar method to transform your data.

Once the preprocessing is completed, partition the data randomly into training and testing sets using a $6$:$4$ ratio. Use the training set to fit a model and the testing set to assess the model performance.

Develop a model using various techniques you$$ve learned from the lecture, and propose and explain the best model.

There are three numeric variables: credit duration month, amount, and age. Plot the histogram of the variables. $(20$ pts$)$

Normalize the above variables, plot the histogram of the variables again. $(20$ pts$)$

Normalize and factorize the appropriate variables, and split in a $6$:$4$ ratio for train:test. Use the training set, and create svm model using e$1071$ package svm$.$ Test the generated model on the test dataset, and explain the results of trained model and the comparison of the original dataset and tested model using pred $(20$ pts$).$

Repeat $($c$)$ but with $$linear kernel$$ and $$non$-$linear kernel$(40$ pts$).$