Discuss your key findings:

1. Did dimensionality reduction improve performance or interpretation?
2. Which classifier performed best and why?
3. What did the clusters reveal about your data?
4. Were there any surprises or inconsistencies in the results?

Objective 3: Feature Importance Question: What are the most significant features affecting the prediction of students' end-of-term grades, and how do they impact the classification results? 1. Dimensionality Reduction: PCA V Technique Name: Principal Component Analysis (PCA) Why Chosen: PCA reduces dimensionality by transforming the original variables into a new set of uncorrelated variables, ordered by the amount of original variance they explain. This helps visualize data and identify potentially important features. from sklearn . decomposition import PCA import matplotlib. pyplot as plt import pandas as pd features = df . drop(columns=[ 'studentid' , 'grade'], axis=1) labels = df [ 'grade' ] # Apply PCA pca = PCA(n_components=2) pca_result = pca. fit_transform(features) # Plotting pit. figure(figsize=(8, 6)) pit. scatter(pca_result[:, 0], pca_result[:, 1], c=labels, cmap='viridis' ) pit. title('PCA of Student Data' ) pit. xlabel( 'Principal Component 1' ) pit. ylabel('Principal Component 2' ) pit . colorbar () pit . show( )PCA of Student Data - 1.5 1.0 2 0.5 0 Principal Component 2 0.0 -2 -0.5 -4 -1.0 -4 -3 -2 -1 0 1 2 W 4 Principal Component 1from sklearn . cluster import KMeans from sklearn . preprocessing import StandardScaler # Standardize features scaler = StandardScaler() features_scaled = scaler . fit_transform(features) # Apply K-Means kmeans = KMeans (n_clusters=3, random_state=42) clusters = kmeans. fit_predict(features_scaled) # Plot clusters pit. figure(figsize=(8, 6)) pit. scatter(pca_result[ :, 0], pca_result[:, 1], c=clusters, cmap="viridis' ) plt. title("K-Means Clustering on PCA-reduced Data" ) pit. xlabel('Principal Component 1' ) pit. ylabel('Principal Component 2" ) pit. colorbar (label='Cluster" ) pit . show( )K-Means Clustering on PCA-reduced Data 2.00 - 1.75 1.50 2 1.25 0 Principal Component 2 1.00 Cluster 0.75 -2 0.50 -4 0.25 0.00 -4 -3 -2 -1 0 1 2 3 4 Principal Component 1