| 1. | Title | Title | Modelling silvicultural thinning using unsupervised machine learning: Evidence from Pterocarpus santalinus trial plots of Odisha, India |
|---|
| 2. | Creator | Author / affiliation / country | Sanjay Saini; Department of Silviculture and Agroforestry, College of Forestry, Odisha University of Agriculture and Technology, Bhubaneswar, Odisha, India; India |
|---|
| 2. | Creator | Author / affiliation / country | Swayam Mallick; Conservator of Forests and State Silviculturist, Department of Forests, Environment and Climate Change, Government of Odisha, Bhubaneswar, Odisha, India; India |
|---|
| 3. | Subject | Keywords | K-means clustering; Hierarchical clustering; Pterocarpus santalinus; Silvicultural thinning; Silhouette score; Caliński–Harabasz index |
|---|
| 3. | Subject | Discipline(s) | Silviculture; Forest management; Machine learning |
|---|
| 4. | Description | Abstract | Silvicultural thinning is essential for regulating stand density, yet conventional prescriptions rely on fixed rules and subjective judgment that may not fully account for stand heterogeneity. This study applies unsupervised machine learning to support silvicultural thinning decisions in Pterocarpus santalinus trial plots across Odisha, India. Tree-level girth and height data from six plots were analysed using k-means and hierarchical clustering to classify trees for retention or felling. Across all six plots, k-means produced better-separated clusters than hierarchical clustering, with higher silhouette scores (0.53–0.62 versus 0.46–0.61) and higher Caliński–Harabasz index values (approximately 538–2504 versus 200–2334). The clustering approach offers an objective, reproducible, and quantitative framework for characterising intra-stand variability that can complement traditional rule-based thinning and support more consistent, data-informed silvicultural management. |
|---|
| 5. | Publisher | Publisher | Contemporary Journal Concept Press |
|---|
| 6. | Contributor | Contributors / supporting agencies | Department of Forests; Environment and Climate Change; Government of Odisha; India |
|---|
| 7. | Date | Publication date | 2026-10-30 |
|---|
| 8. | Type | Type | Text.Serial.Journal; Unsupervised machine learning; k-means clustering; hierarchical clustering (average linkage); silhouette analysis; Caliński–Harabasz index |
|---|
| 9. | Format | Format | PDF; application/pdf |
|---|
| 10. | Identifier | Article URL | https://mcfns.com/index.php/Journal/en/article/view/330 |
|---|
| 11. | Source | Journal / volume / issue / year / pages | Mathematical and Computational Forestry & Natural-Resource Sciences (MCFNS); Vol. 18 No. 2: MCFNS October 30, 2026; 59-68 (10) |
|---|
| 11. | Source | Online ISSN | 1946-7664 |
|---|
| 12. | Language | Language | en |
|---|
| 13. | Relation | Supplementary files | |
|---|
| 14. | Coverage | Geographical coverage | |
|---|
| 15. | Rights | Rights / copyright | Mathematical and Computational Forestry & Natural-Resource Sciences (MCFNS) 2026 |
|---|