Image Details

Choose export citation format:

Ultracool dwarf Science with MachIne LEarning (USMILE). I. Scalable Tree-based Models for Photometric Spectral Classification and New Discoveries from LSST Data Preview 1 and Euclid Quick Data Release 1

  • Authors: Zhoujian Zhang, 周健 张, Yanxia Li, 燕侠 李

Zhoujian Zhang et al 2025 The Astronomical Journal 170 .

  • Provider: AAS Journals

Caption: Figure 7.

Visualization of the 30th decision tree from a USMILE customized classifier; each classifier contains 100 decision trees, with 400 classifiers trained in total. Each split evaluates a single color feature against a threshold, branching into two possible paths: a default direction (blue) and an alternative direction (red). Two example regions of the tree are shown, where “iy,” “zy,” “yj,” “yh,” “yk,” “yw1,” and “yw2” correspond to the iLSST − yLSST, zLSST − yLSST, yLSST − JVHS, yLSST − HVHS, yLSST − KS,VHS, yLSST − W1CatWISE, and yLSST − W2CatWISE colors, respectively. The resulting log-odds scores of the leaf nodes are then combined to yield the final classification probability.

Other Images in This Article

Show More

Copyright and Terms & Conditions

Additional terms of reuse