Data-Efficient Deep Learning: Empirical Guidelines for Training Set Size Estimation in Inertial Sensor Classification

2026-07-10

AI Summary

This study addresses the lack of data-driven methods for estimating required dataset sizes in inertial sensor-based classification tasks, which often leads to unnecessarily high data collection costs. By systematically analyzing learning curve convergence patterns in both binary and multiclass settings, the work reveals—for the first time—that classification accuracy exhibits a stable logarithmic growth with increasing data volume. The authors introduce a novel “quantitative stability point” metric that enables reliable extrapolation of total data requirements from small-scale pilot studies. Validated through deep learning models, logarithmic curve fitting, and cross-dataset evaluation across six real-world datasets (totaling 102.7 hours of recordings), the proposed approach demonstrates that practical performance stability can be achieved with substantially less data than conventional heuristic guidelines suggest.

Abstract

Deep learning models dependency on large-scale inertial datasets presents a significant bottleneck in inertial sensor-based classification tasks, such as human activity recognition and smartphone location recognition. In these domains, data collection requires massive recording campaigns that are complex, time-consuming, and difficult to scale. Currently, data-driven guidelines for determining the minimum sample size required to reach a desired accuracy level do not exist. To address this gap, this study presents a systematic empirical evaluation of learning curve convergence rates in inertial classification. We introduce a unified framework that analyzes classification performance under both binary and multi-class scenarios, and derive an empirical formula to estimate performance relative to dataset size. Testing across six diverse, real-world datasets totaling 102.7 hours of inertial measurements demonstrates that accuracy follows a consistent logarithmic growth pattern, regardless of task complexity. Leveraging this finding, we propose a quantitative stability point metric, defined as the sample size required for the learning curve to stabilize within a predefined mean absolute percentage deviation of its asymptotic maximum. Our analysis reveals that models often reach practical stability with substantially fewer samples than traditional heuristics suggest. Ultimately, we offer a generalizable framework to extrapolate total data requirements from small-scale pilot studies, optimizing the tradeoff between recording effort and model reliability. These findings shift the prevailing paradigm from maximizing data volume toward optimizing data efficiency, offering concrete, data-backed guidelines for planning recording campaigns in inertial sensing applications.

Problem

• data-efficient deep learning
• inertial sensor classification
• training set size estimation
• human activity recognition
• learning curve convergence

Innovation

• data-efficient deep learning
• learning curve convergence
• inertial sensor classification
• sample size estimation
• stability point metric