Empirical Ablation and Ensemble Optimization of a Convolutional Neural Network for CIFAR-10 Classification
For practitioners optimizing CNNs on small-image benchmarks, this ablation study provides a practical guide to which modifications matter, but the findings are incremental and dataset-specific.
This paper empirically ablates 17 modifications to a CNN for CIFAR-10, finding that careful selection of training and architectural changes yields 89.23% accuracy, while indiscriminate depth increases can hurt performance.
Convolutional neural networks (CNNs) remain a central approach in image classification, but their performance depends strongly on architectural and training choices. This paper presents an empirical ablation-based study of CNN optimization for the CIFAR-10 benchmark. The study evaluates 17 progressive modifications involving training duration, learning-rate scheduling, dropout configuration, pooling strategy, network depth, filter arrangement, and dense-layer design. The goal is to identify which changes improve generalization and which increase complexity without improving performance. The baseline model achieved 79.5\% test accuracy. Extending training duration improved performance steadily, whereas several structural redesigns reduced accuracy despite greater architectural variation. Based on the strongest individual configurations, a weighted ensemble was constructed, achieving 86.38\% accuracy in the reduced-data setting and 89.23\% when trained using the full CIFAR-10 dataset. These results suggest that performance gains in CNN-based classification depend less on indiscriminate increases in depth or parameter count than on careful empirical selection of training and architectural modifications. The study therefore highlights the practical value of ablation-oriented optimization and ensemble learning for small-image classification.