Files to Edit and Submit: You will need to edit and submit $($DACs$\_$classification.py$,$

evaluation.py$)$ files

to implement your model for the pulsar dataset. You can copy and paste all the necessary pieces of code

that we wrote in the class. Once you have completed the implementation of your classifier, you should save

both the SVM class#Step $1$:

# Import libraries

# In this section, you can use a search engine to look for the functions that will help you implement the following steps

#Step $2$:

# Load dataset and show basic statistics

# $1.$ Show dataset size $($dimensions$)$

# $2.$ Show what column names exist for the $49$ attributes in the dataset

# $3.$ Show the distribution of the target class CES $4\mathrm{.}0$ Percentile Range column

# $4.$ Show the percentage distribution of the target class CES $4\mathrm{.}0$ Percentile Range column

# Step $3$:

#Clean the dataset $-$ you can eitherhandle the missing values in the dataset

# with the mean of the columns attributes or remove rows the have missing values.

# Step $4$:

#Encode the Categorical Variables $-$ Using OrdinalEncoder from the category$\_$encoders library to encode categorical variables as ordinal integers

# Step $5$:

# Separate predictor variables from the target variable $($attributes $($X$)$ and target variable $($y$)$ as we did in the class$)$

# Create train and test splits for model development. Use the $90\%$ and $20\%$ split ratio

# Use stratifying $($stratify$=$y$)$ to ensure class balance in train$/$test splits

# Name them as X$\_$train, X$\_$test, y$\_$train, and y$\_$test

# Name them as X$\_$train, X$\_$test, y$\_$train, and y$\_$test

X$\_$train $=[]$ # Remove this line after implementing train test split

X$\_$test $=[]$ # Remove this line after implementing train test split

# Do not do steps $6-8$ for the Ramdom Forest Model

# Step $6$:

# Standardize the features $($Import StandardScaler here$)$

# Step $7$:

# Below is the code to convert X$\_$train and X$\_$test into data frames for the next steps

cols $=$ X$\_$train.columns

X$\_$train $=$ pd$.$DataFrame$($X$\_$train, columns$=[$cols$])$ # pd is the imported pandas lirary $-$ Import pandas as pd

X$\_$test $=$ pd$.$DataFrame$($X$\_$test, columns$=[$cols$])$ # pd is the imported pandas lirary $-$ Import pandas as pd

# Step $8-$ Build and train the SVM classifier

# Train SVM with the following parameters. $($use the parameters with the highest accuracy for the model$)$

# $1.$ RBF kernel

# $2.$ C$=10\mathrm{.}0($Higher value of C means fewer outliers$)$

# $3.$ gamma $=0\mathrm{.}3($Linear$)$

# Test the above developed SVC on unseen pulsar dataset samples

# compute and print accuracy score

# Save your SVC model $($whatever name you have given your model$)$ as $.$sav to upload with your submission

# You can use the library pickle to save and load your model for this assignment

# Optional: You can print test results of your model here if you want. Otherwise implement them in evaluation.py file

# Get and print confusion matrix

cm $=[[]]$

# Below are the metrics for computing classification accuracy, precision, recall and specificity

TP $=$ cm$[0,0]$

TN $=$ cm$[1,1]$

FP $=$ cm$[0,1]$

FN $=$ cm$[1,0]$

# Compute Precision and use the following line to print it

precision $=0$ # Change this line to implement Precision formula

print$(\text{'}$Precision : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($precision$))$

# Compute Recall and use the following line to print it

recall $=0$ # Change this line to implement Recall formula

print$(\text{'}$Recall or Sensitivity : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($recall$))$

# Compute Specificity and use the following line to print it

specificity $=0$ # Change this line to implement Specificity formula

print$(\text{'}$Specificity : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($specificity$))$

# Step $9$: Build and train the Random Forest classifier

# Train Random Forest with the following parameters.

# $($n$\_$estimators$=10,$ random$\_$state$=0)$

# Test the above developed Random Forest model on unseen DACs dataset samples

# compute and print accuracy score

# Save your Random Forest model $($whatever name you have given your model$)$ as $.$sav to upload with your submission

# You can use the library pickle to save and load your model for this assignment

# Optional: You can print test results of your model here if you want. Otherwise implement them in evaluation.py file

# Get and print confusion matrix

cm $=[[]]$

# Below are the metrics for computing classification accuracy, precision, recall and specificity

TP $=$ cm$[0,0]$

TN $=$ cm$[1,1]$

FP $=$ cm$[0,1]$

FN $=$ cm$[1,0]$

# Compute Classification Accuracy and use the following line to print it

classification$\_$accuracy $=0$

print$(\text{'}$Classification accuracy : $\{0$:$0\mathrm{.}4$f$\}\text{'}.$format$($classification$\_$accuracy$))$

# Compute Precision and use the following line to print it

precision $=0$ # Change this line to implement Precision formula

print$(\text{'}$Precision : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($precision$))$

# Compute Recall and use the following line to print it

recall $=0$ # Change this line to implement Recall formula

print$(\text{'}$Recall or Sensitivity : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($recall$))$

# Compute Specificity and use the following line to print it

specificity $=0$ # Change this line to implement Specificity formula

print$(\text{'}$Specificity : $\{0$:$0\mathrm{.}3$f$\}\text{'}.$format$($specificity$))$