Self-Healing Neural Networks: A path towards Resilient and Autonomous AI Systems

Read full text Summarise with VIRAT
Authors: Thakur Rohith Singh, M. Venkatesh, Boini Sravan Kumar, Shravya Sujit Kamble
Cited by: 0
References: 6

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Self-Healing Neural Networks (SHNNs) represent an advanced approach in Artificial Intelligence designed to improve system resilience and autonomy. Traditional neural networks, though highly accurate, are vulnerable to hardware faults, adversarial attacks, and data drift in mission-critical environments. The proposed SHNN framework integrates fault detection, adaptive retraining, redundancy, and meta-learning to enable automatic recovery without human intervention. By continuously monitoring internal activations and performance metrics, the system identifies degradation and applies corrective measures such as selective retraining and weight reallocation. Experimental results show that SHNNs maintain high accuracy and system uptime even under fault conditions, enhancing overall reliability and robustness. Keywords: Self-Healing Neural Networks (SHNN), Resilient AI, Fault Detection, Adaptive Learning, Meta-Learning, Autonomous Recovery, Neural Network Robustness, Sustainable AI Systems.

Fields of Study & Subjects

Neural Networks and Applications Adversarial Robustness in Machine Learning Fault Detection and Control Systems

Publication details

Journal / source
INTERANTIONAL JOURNAL OF SCIENTIFIC RESEARCH IN ENGINEERING AND MANAGEMENT
Volume
10
Issue
03
Pages
1–9
Publication date
31 March 2026
Language
EN