Improving Convolutional Neural Network Design via Variable Neighborhood Search

• Type: Article in International Conference Proceedings Book
• Date: 2017

Authors

Araujo, T (Author) Other; Aresta, G (Author) Other; Bernardo Almada Lobo (Author) FEUP; Ana Maria Mendonça (Author) FEUP; Aurélio Campilho (Author) FEUP

Conference proceedings International

Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Pages: 371-379

14th International Conference on Image Analysis and Recognition, ICIAR 2017

5 July 2017 through 7 July 2017

Indexing

ISI Web of Knowledge - 1 Citation

Scopus - 1 Citation

Other information

Authenticus ID: P-00M-WJ8

DOI: 10.1007/978-3-319-59876-5_41

Abstract (EN): An unsupervised method for convolutional neural network (CNN) architecture design is proposed. The method relies on a variable neighborhood search-based approach for finding CNN architectures and hyperparameter values that improve classification performance. For this purpose, t-Distributed Stochastic Neighbor Embedding (t-SNE) is applied to effectively represent the solution space in 2D. Then, k-Means clustering divides this representation space having in account the relative distance between neighbors. The algorithm is tested in the CIFAR-10 image dataset. The obtained solution improves the CNN validation loss by over 15% and the respective accuracy by 5%. Moreover, the network shows higher predictive power and robustness, validating our method for the optimization of CNN design. © Springer International Publishing AG 2017.

Language: English

Type (Professor's evaluation): Scientific