Answer: Unsupervised Learning (Clustering)

## Explanation **Unsupervised Learning (Clustering)** is the correct choice because: 1. **No labeled data available**: The problem states "without having topic labels," which means there are no pre-defined categories or labels for the news articles. 2. **Clustering purpose**: Clustering algorithms are specifically designed to group similar data points together based on their features without any prior knowledge of categories. 3. **Common use case**: Topic modeling and document clustering are classic applications of unsupervised learning where documents are grouped based on word frequency, TF-IDF scores, or embeddings. **Why other options are incorrect:** - **A) Supervised Learning**: Requires labeled training data with known categories, which is not available in this scenario. - **B) Reinforcement Learning**: Used for decision-making problems where an agent learns through rewards/penalties, not suitable for document grouping. - **D) Transfer Learning**: Involves using knowledge from one domain to improve learning in another domain, but still typically requires some labeled data. **Common clustering algorithms** that could be used for this task include: - K-means clustering - Hierarchical clustering - DBSCAN - Latent Dirichlet Allocation (LDA) for topic modeling The key distinction is that unsupervised learning discovers patterns and structures in data without any guidance from labeled examples.

Author: Jin H