Question 3 (0.3 points) A team is visualizing clusters of high-dimensional e-commerce customer behavior data to recommend personalized products. Which technique is best for creating a 2D plot that preserves local structure? Naive Bayes PCA O t-SNE Question 4 (0.3 points) You're developing a predictive inventory model for a warehouse company. Before modeling, which technique would help you extract the most informative patterns while minimizing computational cost? OK-Means Clustering PCA after standardization Feature Selection with forward selection