QuadraNet V2: Efficient and Sustainable Training of High-Order Neural Networks with Quadratic Adaptation

TL;DR

QuadraNet V2 introduces a quadratic neural network framework that efficiently leverages pre-trained weights to significantly reduce training time while enhancing the modeling of data non-linearity and shifts.

Contribution

It presents a novel quadratic neural network architecture that combines pre-trained primary terms with quadratic adaptation, improving efficiency and modeling capacity.

Findings

Reduces GPU hours by up to 98.4% compared to training from scratch.

Enhances high-order model capacity with quadratic adaptation.

Demonstrates effective transfer learning with pre-trained weights.

Abstract

Machine learning is evolving towards high-order models that necessitate pre-training on extensive datasets, a process associated with significant overheads. Traditional models, despite having pre-trained weights, are becoming obsolete due to architectural differences that obstruct the effective transfer and initialization of these weights. To address these challenges, we introduce a novel framework, QuadraNet V2, which leverages quadratic neural networks to create efficient and sustainable high-order learning models. Our method initializes the primary term of the quadratic neuron using a standard neural network, while the quadratic term is employed to adaptively enhance the learning of data non-linearity or shifts. This integration of pre-trained primary terms with quadratic terms, which possess advanced modeling capabilities, significantly augments the information characterization capacity of the high-order network. By utilizing existing pre-trained weights, QuadraNet V2 reduces the required GPU hours for training by 90\% to 98.4\% compared to training from scratch, demonstrating both efficiency and effectiveness.