{"slug": "loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi", "title": "Loss-Aware Feature-Map Pruning in Convolutional Neural Networks Using Multi-Armed Bandits", "summary": "Researchers propose a loss-aware feature-map pruning framework for convolutional neural networks that uses multi-armed bandits (UCB1 and Thompson Sampling) to identify and remove redundant feature maps while preserving accuracy. Evaluated on datasets including MNIST, CIFAR-10, CIFAR-100, SVHN, CUB-200-2011, and Oxford Flowers 102, UCB1 achieved the highest mean rank in Friedman and Nemenyi tests, outperforming greedy and magnitude-based pruning and remaining statistically comparable to unpruned models.", "body_md": "arXiv:2607.22564v1 Announce Type: new\nAbstract: Convolutional neural networks often contain redundant feature maps that increase storage and inference cost. This paper presents a loss-aware feature-map pruning framework using multi-armed bandits. Feature-map pruning is structured because it removes complete convolutional output channels and their producing filters rather than isolated scalar weights. Each candidate feature map is treated as an arm. At each play time, one map is temporarily masked and evaluated on a sampled mini-batch; the map is then restored and the observed loss change is converted into a safe-removal reward. After a fixed play budget, candidate maps are ranked by learned scores and the top-k maps are permanently removed with their filters, biases and corresponding next-layer input-channel kernels. The study evaluates UCB1 and Thompson Sampling, compares them with direct/oracle-style evaluation on LeNet/MNIST, and extends the evaluation to MNIST, CIFAR-10, CIFAR-100, SVHN, CUB-200-2011 and Oxford Flowers 102. Results show that UCB1 and Thompson Sampling preserve accuracy close to unpruned models while removing feature maps and reducing convolutional computation. Friedman and Nemenyi tests show that UCB1 obtains the highest mean rank, followed by Thompson Sampling; both significantly outperform greedy and magnitude-based pruning while remaining statistically comparable to the original unpruned model.", "url": "https://wpnews.pro/news/loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi", "canonical_source": "https://arxiv.org/abs/2607.22564", "published_at": "2026-07-28 04:00:00+00:00", "updated_at": "2026-07-28 04:29:12.380080+00:00", "lang": "en", "topics": ["machine-learning", "neural-networks", "computer-vision"], "entities": ["UCB1", "Thompson Sampling", "LeNet", "MNIST", "CIFAR-10", "CIFAR-100", "SVHN", "CUB-200-2011"], "alternates": {"html": "https://wpnews.pro/news/loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi", "markdown": "https://wpnews.pro/news/loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi.md", "text": "https://wpnews.pro/news/loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi.txt", "jsonld": "https://wpnews.pro/news/loss-aware-feature-map-pruning-in-convolutional-neural-networks-using-multi.jsonld"}}