Advancing Taxonomy with Machine Learning: A Hybrid Ensemble for Species and Genus Classification

Algorithms, Journal Year: 2025, Volume and Issue: 18(2), P. 105 - 105

Published: Feb. 14, 2025

Traditionally, classifying species has required taxonomic experts to carefully examine unique physical characteristics, a time-intensive and complex process. Machine learning offers promising alternative by utilizing computational power detect subtle distinctions more quickly accurately. This technology can classify both known (described) unknown (undescribed) species, assigning samples specific grouping ones at the genus level—an improvement over common practice of labeling as outliers. In this paper, we propose novel ensemble approach that integrates neural networks with support vector machines (SVM). Each animal is represented an image its DNA barcode. Our research investigates transformation one-dimensional data into two-dimensional three-channel matrices using discrete wavelet transform (DWT), enabling application convolutional (CNNs) have been pre-trained on large datasets. method significantly outperforms existing approaches, demonstrated several datasets containing images barcodes. By classification described undescribed represents major step forward in global biodiversity monitoring.

Language: Английский

---

Charting a future for entomological taxonomy in New Zealand

New Zealand Entomologist, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 17

Published: Sept. 25, 2024

Language: Английский

---