A theoretical framework proposes two linear equations for predicting inference latency and energy across binary classifiers, but no coefficients are calibrated and no validation is provided.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Latenrgy: Model Agnostic Latency and Energy Consumption Prediction for Binary Classifiers
A theoretical framework proposes two linear equations for predicting inference latency and energy across binary classifiers, but no coefficients are calibrated and no validation is provided.