QIBONN: A Quantum-Inspired Bilevel Optimizer for Neural Networks on Tabular Classification

2025-11-12

AI Summary

Addressing the challenges of large search spaces, strong non-convexity, and high tuning costs in neural hyperparameter optimization (HPO) for tabular data, this paper proposes QIBONN, a quantum-inspired bilevel optimization framework. Methodologically, QIBONN employs qubit-based unified encoding to jointly represent feature selection, network architecture, and regularization hyperparameters; introduces a hybrid search mechanism integrating deterministic rotation with global-attractor-guided stochastic mutation to balance exploration and exploitation; and incorporates IBM-Q–inspired single-qubit flip noise modeling to enhance robustness. Extensive experiments across 13 real-world tabular datasets demonstrate that, under identical tuning budgets, QIBONN matches or surpasses state-of-the-art tree-based models and existing classical/quantum-inspired HPO methods in predictive performance. The framework establishes a novel, efficient, and scalable paradigm for HPO in tabular learning.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Quantum Machine Learning
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Hyperparameter optimization (HPO) for neural networks on tabular data is critical to a wide range of applications, yet it remains challenging due to large, non-convex search spaces and the cost of exhaustive tuning. We introduce the Quantum-Inspired Bilevel Optimizer for Neural Networks (QIBONN), a bilevel framework that encodes feature selection, architectural hyperparameters, and regularization in a unified qubit-based representation. By combining deterministic quantum-inspired rotations with stochastic qubit mutations guided by a global attractor, QIBONN balances exploration and exploitation under a fixed evaluation budget. We conduct systematic experiments under single-qubit bit-flip noise (0.1%--1%) emulated by an IBM-Q backend. Results on 13 real-world datasets indicate that QIBONN is competitive with established methods, including classical tree-based methods and both classical/quantum-inspired HPO algorithms under the same tuning budget.

Problem

• Optimizing neural network hyperparameters for tabular classification tasks
• Addressing large non-convex search spaces in hyperparameter optimization
• Reducing computational costs of exhaustive tuning through quantum-inspired methods

Innovation

• Quantum-inspired bilevel optimizer for neural networks
• Unified qubit representation for feature and architecture selection
• Combines deterministic rotations with stochastic qubit mutations