A softmax-sampled, budgeted replacement for the classical sequential search step keeps per-step cost independent of dimensionality and, with online-learned feature statistics, matches or beats ranking baselines up to 10^4 features.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Stochastic Sequential Search in Very-High-Dimensional Feature Selection
A softmax-sampled, budgeted replacement for the classical sequential search step keeps per-step cost independent of dimensionality and, with online-learned feature statistics, matches or beats ranking baselines up to 10^4 features.