Today, convolutional neural network (CNN) pruning techniques often rely on manually crafted importance criteria and pruning structures. Due to their heuristic nature, these methods may lack generality, and their performance is not guaranteed. In this paper, we propose a theoretical framework to address this challenge by leveraging the concept of $\gamma$γ-weak submodularity, based on a new efficient importance function. By deriving an upper bound on the absolute error in the layer subsequent to the pruned layer, we formulate the importance function as a $\gamma$γ-weakly submodular function. This formulation enables the development of an easy-to-implement, low-complexity, and data-free oblivious algorithm for selecting filters to be removed from a convolutional layer. Extensive experiments show that our method outperforms state-of-the-art benchmark networks across various datasets, with a computational cost comparable to the simplest pruning techniques, such as $l_{2}$l2-norm pruning. Notably, the proposed method achieves an accuracy of 76.52%, compared to 75.15% for the overall best baseline, with a 25.5% reduction in network parameters. According to our proposed resource-efficiency metric for pruning methods, the ACLI approach demonstrates orders-of-magnitude higher efficiency than the other baselines, while maintaining competitive accuracy.

Convolutional neural network (CNN) pruning has traditionally relied on heuristically designed importance criteria, often leading to limited generalizability and inconsistent performance. In this article, we propose a novel framework centered around filter replacement (FR), introducing pruning as a process of replacing selected filters with zero filters. Through a rigorous analysis, we derive an upper bound on the absolute error in the output of the subsequent layer and use this bound to define an efficient importance function. This importance function exhibits $\gamma $ -weakly submodular properties, enabling the development of a simple, low-complexity, and data-free oblivious algorithm for selecting filters to prune. In addition, we extend the FR framework to include nonzero filter alternatives, leveraging a best-approximation technique to construct optimal replacements for the pruned filters. Extensive experiments on benchmark networks and datasets validate the effectiveness of our method. The proposed approach achieves state-of-the-art results, with a complexity comparable to basic techniques such as $l_{2}$ -norm pruning. Notably, our pruning method achieves 76.52% accuracy (ACC) in ResNet-50 on the ImageNet dataset, surpassing the baseline of 75.15%, while reducing network parameters by 25.5%. Our proposed resource efficiency (RE) metric assesses that the layer interdependence-aware pruning (LIAP) method is up to $10^{11}$ times more efficient than existing techniques, setting a new standard for resource-aware CNN pruning.

Today’s machine learning applications in the healthcare sector often require a vast amount of training data and lengthy training, which are not always available. Transfer learning (TL) offers a promising solution by leveraging knowledge across different domains/tasks, effectively addressing data scarcity. However, TL encounters computational and communication challenges in resource-constrained scenarios, and negative transfer (NT) can arise, posing significant challenges for sensitive tasks such as stroke prediction. This paper presents a novel method to maximize the accuracy of TL in resource-constrained environments while mitigating NT, a key concern in TL. Previous studies have overlooked simultaneously the resource consumption and the NT problem in the network-based TL era. To address this shortcoming, we formulate an optimization model named RCTL, which employs the empirical fine-tuning error of the target task as the objective while considering NT overcoming condition and resource limitations as the feasibility constraints. Our approach is extensively validated in a stroke prediction scenario, demonstrating the efficiency of RCTL in achieving the desired trade-offs between TL’s benefits and associated costs. The obtained results show the potential of enhancing edge AI applications in resource-constrained healthcare infrastructure.

ThiNet is a recent method for pruning convolutional neural networks. This method uses a norm of a subset of the components of the output resulting from the convolutional layer succeeding the layer from which the filters are to be removed for pruning the network. The ThiNet algorithm is very time-consuming, in view of the fact that the filters for removal are selected one by one iteratively. In this paper, we propose a modified version of ThiNet, in which the same information on the output of the same convolutional layer as used by ThiNet is employed to select all the filters together in a single step, for pruning the network. The proposed modified algorithm is shown to have a time-complexity that is only a small fraction of that of ThiNet or any other state-of-the-art algorithm and that the pruned network has almost the same reduction in its accuracy as that of the network pruned by ThiNet.

In this study, the authors present a new criterion which can be used to decompose the coding advantage (CA) of space–frequency block codes (SFBCs). The proposed decomposition separates the CA into two parts, called precoder‐CA and channel‐CA. The precoder‐CA represents the effects of the precoder and the signal constellation, and can be used in the design of the precoder of SFBCs. The channel‐CA, on the other hand, gives an indication of the effects of the channel characteristics, and can be utilised for optimising the permutation parameter. The authors have shown that the proposed decomposition method leads to a lower computational complexity for the optimisation procedure compared with the general form of the CA. As an application of the proposed method, the authors used it to optimise the channel‐CA of a high‐rate SFBC. Simulation results confirmed better performance for the modified high‐rate SFBC compared with recently proposed SFBCs in the literature.