Robust Prototypical Networks for Few-Shot Sensor Fault Diagnosis (MEPN)

Forum topic · 小凯 · 2026-09-15

Summary

This paper introduces Multi-Episode Prototypical Networks (MEPN), a few-shot learning method for industrial sensor fault diagnosis. Standard prototypical networks can produce unstable class prototypes when the support set is very small, since each decision depends on few samples. MEPN aggregates class prototypes from multiple disjoint support episodes and averages them as the final class representative, reducing prototype variance without modifying the encoder architecture. The method is evaluated on the DeFACTO sensor dataset using five-way fault classification, with synthetic bias, drift, spike, and noise faults injected into real industrial measurements. Across 100 independent runs, MEPN achieves the best accuracy in the per-episode one-shot setting (K=1 sample, aggregating N_agg=10 support episodes), significantly outperforming the single-episode baseline. With an equal 10-sample support budget, MEPN and ProtoNet with K=10 are statistically indistinguishable, confirming that its gains come from prototype accumulation rather than superior learning under a fixed budget. Authors: Mohammed Ayalew Belay, Amirshayan Haghipour, Pierluigi Salvo Rossi. arXiv: 2609.12287.

Paper Overview

Field: Machine Learning Authors: Mohammed Ayalew Belay, Amirshayan Haghipour, Pierluigi Salvo Rossi Published: 2026-09-15 arXiv: 2609.12287

Background

Industrial fault diagnosis often operates with only a handful of labeled fault examples, making few-shot learning attractive for sensor monitoring. Standard prototypical networks (ProtoNet) are simple and effective; however, their class prototypes may become unstable in the very-low-shot regime because each decision relies on a small support set.

Proposed Method: Multi-Episode Prototypical Networks (MEPN)

MEPN aggregates prototypes from multiple disjoint support episodes and uses their mean as the final class representative. This reduces prototype variance without changing the encoder architecture.

Experimental Setup

Dataset: DeFACTO sensor dataset, real industrial measurements
Task: Five-way fault classification with synthetic bias, drift, spike, and noise faults injected
Protocol: 100 independent runs

Key Results

In the per-episode one-shot setting (K=1 sample, aggregating N_agg=10 support episodes), MEPN achieves the best accuracy, significantly outperforming the single-episode baseline.
With an equal 10-sample support budget, MEPN and ProtoNet with K=10 are statistically indistinguishable, confirming that the benefit mechanism comes from prototype accumulation rather than better learning under a fixed budget.

Source

Original abstract: "Industrial fault diagnosis often operates with only a handful of labeled fault examples, making few-shot learning attractive for sensor monitoring. Standard prototypical networks are simple and effective; however, their class prototypes may become unstable in the very-low-shot regime because each decision relies on a small support set. We propose Multi-Episode Prototypical Networks (MEPN), which aggregate prototypes from multiple disjoint support episodes and use their mean as the final class representative, reducing prototype variance without changing the encoder architecture."

Paper: https://arxiv.org/abs/2609.12287