LTLS

2000년 도입 · 논문 2편에서 사용

LTLS is a technique for multiclass and multilabel prediction that can perform training and inference in logarithmic time and space. LTLS embeds large classification problems into simple structured prediction problems and relies on efficient dynamic programming algorithms for inference. It tackles extreme multi-class and multi-label classification problems where the size $C$ of the output space is extremely large.

출처: Log-time and Log-space Extreme Classification

소개 논문: Log-time and Log-space Extreme Classification