A carregar...
Publicação
Effective and efficient image classification from Deep Features Statistics
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.68 MB | Adobe PDF |
Orientador(es)
Resumo(s)
In this thesis, "Effective and Efficient Image Classification from Deep Features Statistics," it is presented an in-depth analysis of the balance between computational efficiency and the accuracy of image classification models. The study leverages pre-trained deep learning architectures—VGG16, DenseNet121, and ResNet50—on CIFAR-10 and SVHN datasets, aiming to reduce training time without compromising accuracy. The research introduces a novel tradeoff ratio metric, providing a pragmatic approach to evaluate the efficacy of Transfer Learning (T.L.) against traditional Machine Learning (M.L.) techniques. The findings illuminate pathways for time-efficient image classification, making significant contributions to both academic research and practical applications in the field.
Descrição
Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science
Palavras-chave
Artificial Intelligence CNN Statistics Machine Learning Transfer Learning SDG 9 - Industry, innovation and infrastructure SDG 11 - Sustainable cities and communities SDG 12 - Responsible production and consumption SDG 13 - Climate action SDG 15 - Life on land SDG 17 - Partnerships for the goals