En esta sección se presenta información acerca de publicaciones realizadas:
- Journal articles
- Conferences articles
- Book chapters
- Books
- Inproceedings.
Puede seleccionar opciones de visualización: year and / or articles type / all types
2021 |
|
2. | Fernández-Fernández Y; Peluffo-Ordóñez D.H.; Umaquinga-Criollo A.C.; Lorente-Leyva L.L.; Cabrera-Alvarez E.N International Conference on Communication, Computing and Electronics Systems. Proceedings of ICCCES 2020, vol. 733, Springer Science and Business Media Deutschland GmbH, 2021, ISBN: 978-981-33-4908-7. Abstract | Links | BibTeX | Etiquetas: Classification, Nearest neighbor algorithms, Prototypes @conference{fernandez2021_, The condensed nearest neighbor (CNN) classifier is one of the techniques used and known to perform recognition tasks. It has also proven to be one of the most interesting algorithms in the field of data mining despite its simplicity. However, CNN suffers from several drawbacks, such as high storage requirements and low noise tolerance. One of the characteristics of CNN is that it focuses on the selection of prototypes, which consists of reducing the set of training data. One of the goals of CNN seeks to achieve the reduction of information in such a way that the reduced information can represent large amounts of data to exercise decision-making on them. This paper mentions some of the most recent contributions to CNN-based unsupervised algorithms in a review that builds on the mathematical principles of condensed methods. |
2018 |
|
1. | P D Rosero-Montalvo; A C Umaquinga-Criollo; S Flores; L Suarez; J Pijal; K L Ponce-Guevara; D Nejer; A Guzman; D Lugo; K Moncayo Neighborhood Criterion Analysis for Prototype Selection Applied in WSN Data Conference 2017 International Conference on Information Systems and Computer Science (INCISCOS), IEEE, 2018, ISBN: 978-1-5386-2644-3, (Electronic ISBN: 978-1-5386-2644-3 Print on Demand(PoD) ISBN: 978-1-5386-2645-0). Abstract | Links | BibTeX | Etiquetas: {WSN} data, classification and the reduction of data set, Computer science, data reduction, data subset criterion, Information systems, learning (artificial intelligence), Machine learning algorithms, neighborhood criterion analysis, normalized distance, pattern classification, prototype selection, Prototypes, redundant data, set theory, Silicon, supervised machine learning classification algorithms, Training, training matrix, wireless sensor networks @conference{rosero-montalvo_neighborhood_2017, The present work presents an analysis of the neighborhood criterion for the prototype selection (PS) in supervised machine learning classification algorithms. To do this, we use the condensed neighbor algorithm CNN to eliminate redundant data with the normalization of the distance to the centroid of each data subset criterion. This is done, in order to obtain the training matrix of the most optimal model. A selection of neighborhood criterion has been created from the quantification of the balance between the performance of the classification and the reduction of data set (CER). As proof of the test, we performed: (i) CER and (ii) real-time tests with the implementation of the algorithm within the WSN. The result is a data reduction of up to 88 % and a performance of the kNN classifier of 75%. It is concluded that the criterion of neighborhood with normalized distance must be less than or equal to 0.2 and the implementation of kNN with k = 1 obtains the best CER. |