1️⃣ Question 1

Why run an unsupervised model when no diagnoses are available?

• ❌ Rank patients by admission time

• ❌ Compress features into one index

• ❌ Assign diagnosis codes

• ✅ To uncover natural patient groups that share similar vital sign patterns

Explanation:
Unsupervised learning finds patterns without labels, perfect when no diagnosis labels exist.

---

2️⃣ Question 2

How does hierarchical clustering (dendrogram) help determine the number of clusters?

• ✅ Visualizes similarity levels, helping decide an optimal cluster count

• ❌ Produces same clusters every time

• ❌ Removes outliers automatically

• ❌ Designed for high-dimensional data

Explanation:
A dendrogram shows where clusters merge, helping identify a cutoff.

---

3️⃣ Question 3

What does a K-means centroid represent?

• ❌ Range of spending

• ❌ Distance between customers and center

• ❌ Number of customers improving

• ✅ Average spend and purchase frequency of the cluster

Explanation:
A centroid is the mean of all points in a cluster.

---

4️⃣ Question 4

Why is DBSCAN good for detecting unusual activity?

• ❌ Needs exact number of clusters

• ✅ Identifies clusters of various shapes and detects outliers

• ❌ Assigns every user to a cluster

• ❌ Minimizes user-center distances

Explanation:
DBSCAN naturally handles noise and irregular clusters.

---

5️⃣ Question 5

Why is t-SNE used for 2D scatter plots?

• ❌ Auto-labels segments

• ❌ Enforces linear projections

• ❌ Equalizes distances

• ✅ Maintains neighborhood similarities for visual discovery

Explanation:
t-SNE preserves local structure, revealing natural groupings.

---

6️⃣ Question 6

Why is PCA suitable for environmental factor research?

• ❌ Finds nonlinear relationships

• ✅ Reduces data to key components for simpler analysis

• ❌ Combines all variables into one

• ❌ Forces features to be uncorrelated (this is a result, not the purpose)

Explanation:
PCA reduces dimensionality while retaining major variance.

---

7️⃣ Question 7

Advantage of t-SNE for user interaction visualization:

• ❌ Eliminates noise

• ❌ Guarantees linear transformation

• ✅ Allows clusters to appear clearly in low-dimensional space

• ❌ Reduces to a single dimension

Explanation:
t-SNE creates a meaningful lower-dimensional visualization where similar users cluster together.

---

🧾 Summary Table

---