# Evaluate on the test set

y$\_$pred $=$ grid$\_$svm$.$predict$($X$\_$test$)$

print$("$SVM Test Accuracy:", accuracy$\_$score$($y$\_$test, y$\_$pred$))$

print$("$Classification Report:

$",$ classification$\_$report$($y$\_$test, y$\_$pred$))$

print$("$Confusion Matrix:

$",$ confusion$\_$matrix$($y$\_$test, y$\_$pred$))$

# # K$-$Nearest Neighbors $($KNN$)$

# Define hyperparameters for KNN

param$\_$grid$\_$knn $=\{$

'classifier$\_\_$n$\_$neighbors': $[3,5,7,10],$

'classifier$\_\_$weights': $[\text{'}$uniform$\text{'},$ 'distance'$],$

'classifier$\_\_$p$\text{'}$: $[1,2]$

$\}$

# Set up GridSearchCV

grid$\_$knn $=$ GridSearchCV$($Pipeline$([$

$(\text{'}$preprocessor$\text{'},$ preprocessor$),$

$(\text{'}$classifier$\text{'},$ KNeighborsClassifier$())$

$]),$ param$\_$grid$\_$knn$,$ cv$=5,$ scoring$=$'accuracy'$)$

# Fit GridSearchCV

grid$\_$knn$.$fit$($X$\_$train, y$\_$train$)$

# Best parameters and score

print$("$Best Parameters for KNN:$",$ grid$\_$knn$.$best$\_$params$\_)$

print$("$Best Score for KNN:$",$ grid$\_$knn$.$best$\_$score$\_)$

# Evaluate on the test set

y$\_$pred $=$ grid$\_$knn$.$predict$($X$\_$test$)$

print$("$KNN Test Accuracy:", accuracy$\_$score$($y$\_$test, y$\_$pred$))$

print$("$Classification Report:

$",$ classification$\_$report$($y$\_$test, y$\_$pred$))$

print$("$Confusion Matrix:

$",$ confusion$\_$matrix$($y$\_$test, y$\_$pred$))$

# # Train and Evaluate the Models

# Store results in a dictionary

results $=\{\}$

# Logistic Regression

y$\_$pred $=$ grid$\_$log$\_$reg.predict$($X$\_$test$)$

results$[\text{'}$Logistic Regression'$]=\{$

'Accuracy': accuracy$\_$score$($y$\_$test, y$\_$pred$),$

'Classification Report': classification$\_$report$($y$\_$test, y$\_$pred, output$\_$dict$=$True$),$

'Confusion Matrix': confusion$\_$matrix$($y$\_$test, y$\_$pred$)$

$\}$

# Decision Tree

y$\_$pred $=$ grid$\_$dec$\_$tree.predict$($X$\_$test$)$

results$[\text{'}$Decision Tree'$]=\{$

'Accuracy': accuracy$\_$score$($y$\_$test, y$\_$pred$),$

'Classification Report': classification$\_$report$($y$\_$test, y$\_$pred, output$\_$dict$=$True$),$

'Confusion Matrix': confusion$\_$matrix$($y$\_$test, y$\_$pred$)$

$\}$

# Random Forest

y$\_$pred $=$ grid$\_$rf$.$predict$($X$\_$test$)$

results$[\text{'}$Random Forest'$]=\{$

'Accuracy': accuracy$\_$score$($y$\_$test, y$\_$pred$),$

'Classification Report': classification$\_$report$($y$\_$test, y$\_$pred, output$\_$dict$=$True$),$

'Confusion Matrix': confusion$\_$matrix$($y$\_$test, y$\_$pred$)$

$\}$

# SVM

y$\_$pred $=$ grid$\_$svm$.$predict$($X$\_$test$)$

results$[\text{'}$SVM$\text{'}]=\{$

'Accuracy': accuracy$\_$score$($y$\_$test, y$\_$pred$),$

'Classification Report': classification$\_$report$($y$\_$test, y$\_$pred, output$\_$dict$=$True$),$

'Confusion Matrix': confusion$\_$matrix$($y$\_$test, y$\_$pred$)$

$\}$

# KNN

y$\_$pred $=$ grid$\_$knn$.$predict$($X$\_$test$)$

results$[\text{'}$KNN$\text{'}]=\{$

'Accuracy': accuracy$\_$score$($y$\_$test, y$\_$pred$),$

'Classification Report': classification$\_$report$($y$\_$test, y$\_$pred, output$\_$dict$=$True$),$

'Confusion Matrix': confusion$\_$matrix$($y$\_$test, y$\_$pred$)$

$\}$

# # Compare Accuracy

import matplotlib.pyplot as plt

# Extract accuracies

models $=$ list$($results$.$keys$())$

accuracies $=[$results$[$model$][\text{'}$Accuracy$\text{'}]$ for model in models$]$

# Plot

plt$.$figure$($figsize$=(10,6))$

plt$.$bar$($models$,$ accuracies, color$=[\text{'}$blue$\text{'},$ 'green', 'red', 'purple', 'orange'$])$

plt$.$title$(\text{'}$Model Accuracy Comparison'$)$

plt$.$xlabel$(\text{'}$Model$\text{'})$

plt$.$ylabel$(\text{'}$Accuracy$\text{'})$

plt$.$ylim$([0,1])$

plt$.$show$()$

# # Visualize Confusion Matrices

import seaborn as sns

def plot$\_$confusion$\_$matrix$($conf$\_$matrix, title$)$:

plt$.$figure$($figsize$=(8,6))$

sns$.$heatmap$($conf$\_$matrix, annot$=$True, fmt$=\text{'}$d$\text{'},$ cmap$=$'Blues', cbar$=$False$)$

plt$.$title$($title$)$

plt$.$xlabel$(\text{'}$Predicted Label'$)$

plt$.$ylabel$(\text{'}$True Label'$)$

plt$.$show$()$

# Plot confusion matrices for each model

for model, result in results.items$()$:

plot$\_$confusion$\_$matrix$($result$[\text{'}$Confusion Matrix'$],$ f$\text{'}\{$model$\}$ Confusion Matrix'$)$

# # Text$-$Based Application

import joblib

# Save the model

joblib.dump$($grid$\_$rf$.$best$\_$estimator$\_,$ 'random$\_$forest$\_$model.pkl$\text{'})$

Explain the code $?$