D
Diego Furtado Silva
Researcher at Federal University of São Carlos
Publications - 69
Citations - 2051
Diego Furtado Silva is an academic researcher from Federal University of São Carlos. The author has contributed to research in topics: Dynamic time warping & Feature extraction. The author has an hindex of 20, co-authored 65 publications receiving 1533 citations. Previous affiliations of Diego Furtado Silva include University of São Paulo & Spanish National Research Council.
Papers
More filters
How k-nearest neighbor parameters affect its performance
TL;DR: This work provides a discussion and presents empirical evidence of how the main parameters of k-Nearest Neighbor influence its performance, including the number of nearest neighbors, distance function and weighting function.
Proceedings ArticleDOI
SiMPle: assessing music similarity using subsequences joins
TL;DR: This work proposes a simple and efficient representation based on a subsequence similarity join, which may be used in several music information retrieval tasks and applies it to the cover song recognition problem and demonstrates that it is superior to state-of-the-art algorithms.
Proceedings ArticleDOI
Applying Machine Learning and Audio Analysis Techniques to Insect Recognition in Intelligent Traps
Diego Furtado Silva,Vinicius M. A. Souza,Gustavo E. A. P. A. Batista,Eamonn Keogh,Daniel P. W. Ellis +4 more
TL;DR: An intelligent trap that uses a laser sensor to selectively classify and catch insects is presented and it is shown that a binary class classifier that recognizes the mosquito species achieved almost perfect accuracy, assuring the applicability of the proposed intelligent trap.
Proceedings ArticleDOI
Extracting Texture Features for Time Series Classification
TL;DR: This work proposes a method capable of extracting texture features from this graphical representation of recurrence plots, and uses those features to classify time series data, and shows that it outperforms the state-of-the-art methods for time series classification.
Proceedings ArticleDOI
Classification of Data Streams Applied to Insect Recognition: Initial Results
TL;DR: The objective of this paper is to evaluate methods that adapt concept drifts by regularly updating the classification models applied to insect recognition in a data stream by showing in the initial results that the philosophy of inserting and removing examples from the training set are of essential importance.