K means Clustering - tahiri-lab/KMeansPhyloTreesClustering GitHub Wiki

🧠 K-means Clustering

General Definition

K-means is an unsupervised machine learning algorithm used to group data into K clusters.

It aims to minimize the distance between data points and their cluster centers.

---

How it works

1. Choose the number of clusters K
2. Initialize cluster centers
3. Assign each data point to the nearest center
4. Recompute cluster centers
5. Repeat until convergence

---

Objective

Minimize:

Within-cluster variance

---

🌳 K-means in Bioinformatics (Phylogenetic Trees)

In this project, the data points are phylogenetic trees, not numerical vectors.

---

Key adaptation

• Distance between trees = RF distance
• Trees are grouped based on structural similarity

---

Process in this context

1. Compute RF distance between trees
2. Apply K-means using this distance
3. Group trees with similar topology

---

Interpretation

Each cluster represents:

• a group of trees sharing similar evolutionary structures
• a consistent evolutionary hypothesis

---

🔗 RF + K-means Integration

The combination of RF distance and K-means allows:

• comparison of complex tree structures
• grouping of similar evolutionary hypotheses
• reduction of multiple trees into meaningful clusters

---