Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 2 of 2
Full-Text Articles in Computer Engineering
Assessment Of E-Senses Performance Through Machine Learning Models For Colombian Herbal Teas Classification, Jeniffer Katerine Carrillo, Cristhian Manuel Durán, Juan Martin Cáceres, Carlos Alberto Cuastumal, Jordana Ferreira, José Ramos, Brian Bahder, Martin Oates, Antonio Ruiz
Assessment Of E-Senses Performance Through Machine Learning Models For Colombian Herbal Teas Classification, Jeniffer Katerine Carrillo, Cristhian Manuel Durán, Juan Martin Cáceres, Carlos Alberto Cuastumal, Jordana Ferreira, José Ramos, Brian Bahder, Martin Oates, Antonio Ruiz
CCE Faculty Articles
This paper describes different E-Senses systems, such as Electronic Nose, Electronic Tongue, and Electronic Eyes, which were used to build several machine learning models and assess their performance in classifying a variety of Colombian herbal tea brands such as Albahaca, Frutos Verdes, Jaibel, Toronjil, and Toute. To do this, a set of Colombian herbal tea samples were previously acquired from the instruments and processed through multivariate data analysis techniques (principal component analysis and linear discriminant analysis) to feed the support vector machine, K-nearest neighbors, decision trees, naive Bayes, and random forests algorithms. The results of the E-Senses were validated using …
Adversarial Training Of Deep Neural Networks, Anabetsy Termini
Adversarial Training Of Deep Neural Networks, Anabetsy Termini
CCE Theses and Dissertations
Deep neural networks used for image classification are highly susceptible to adversarial attacks. The de facto method to increase adversarial robustness is to train neural networks with a mixture of adversarial images and unperturbed images. However, this method leads to robust overfitting, where the network primarily learns to recognize one specific type of attack used to generate the images while remaining vulnerable to others after training. In this dissertation, we performed a rigorous study to understand whether combinations of state of the art data augmentation methods with Stochastic Weight Averaging improve adversarial robustness and diminish adversarial overfitting across a wide …