On the Segmentation and Classification of Hand Radiographs

Davis, Luke M.; Theobald, Barry-John; Lines, Jason; Toms, Andoni P.; Bagnall, Anthony

doi:10.1142/s0129065712500207

Cited by 20 publications

(5 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We assessed the performance of EuSN on several real-world time-series classification benchmarks of diverse nature. The first 10 datasets were taken from the UEA & UCR time-series classification repository [45], namely: Adiac [46], CharacterTrajectories [47], ECG5000 [48], Epilepsy [49], HandOutlines [50], Hearthbeat [48], Libras [51], ShapesAll [52], SpokenArabicDigits [53], and Wafer [54]. Moreover, we have considered the IMDB movie review sentiment classification dataset [55], and the Reuters newswire classification dataset from UCI [56].…”

Section: Time-series Classificationmentioning

confidence: 99%

Euler State Networks: Non-dissipative Reservoir Computing

Gallicchio¹

2022

Preprint

View full text Add to dashboard Cite

Inspired by the numerical solution of ordinary differential equations, in this paper we propose a novel Reservoir Computing (RC) model, called the Euler State Network (EuSN). The introduced approach makes use of forward Euler discretization and anti-symmetric recurrent matrices to design reservoir dynamics that are both stable and non-dissipative by construction.Our mathematical analysis shows that the resulting model is biased towards unitary effective spectral radius and zero local Lyapunov exponents, intrinsically operating at the edge of stability. Experiments on synthetic tasks indicate the marked superiority of the proposed approach, compared to standard RC models, in tasks requiring longterm memorization skills. Furthermore, results on real-world time series classification benchmarks point out that EuSN is capable of matching (or even surpassing) the level of accuracy of trainable Recurrent Neural Networks, while allowing up to 100-fold savings in computation time and energy consumption.

show abstract

Section: Time-series Classificationmentioning

confidence: 99%

Euler State Networks: Non-dissipative Reservoir Computing

Gallicchio¹

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Given that this is the first time that time series ordinal classification is studied, a subset of 7 time series datasets has been appropriately chosen from the original UCR data repository [28]. Most of the datasets selected come from the field of bone age prediction, presented in [29]. Specifically, those named "AgeGroup" include patterns (bones) labelled as inf ant, junior or teen, depending on the age group to which the bone belongs.…”

Section: A Ordinal Time Series Consideredmentioning

confidence: 99%

Time series ordinal classification via shapelets

Guijo-Rubio

Gutiérrez

Bagnall

et al. 2020

2020 International Joint Conference on Neural Networks (IJCNN)

Self Cite

View full text Add to dashboard Cite

Nominal time series classification has been widely developed over the last years. However, to the best of our knowledge, ordinal classification of time series is an unexplored field, and this paper proposes a first approach in the context of the shapelet transform (ST). For those time series dataset where there is a natural order between the labels and the number of classes is higher than 2, nominal classifiers are not capable of achieving the best results, because the models impose the same cost of misclassification to all the errors, regardless the difference between the predicted and the ground-truth. In this sense, we consider four different evaluation metrics to do so, three of them of an ordinal nature. The first one is the widely known Information Gain (IG), proved to be very competitive for ST methods, whereas the remaining three measures try to boost the order information by refining the quality measure. These three measures are a reformulation of the Fisher score, the Spearman's correlation coefficient (ρ), and finally, the Pearson's correlation coefficient (R 2 ). An empirical evaluation is carried out, considering 7 ordinal datasets from the UEA & UCR time series classification repository, 4 classifiers (2 of them of nominal nature, whereas the other 2 are of ordinal nature) and 2 performance measures (correct classification rate, CCR, and average mean absolute error, AM AE). The results show that, for both performance metrics, the ST quality metric based on R 2 is able to obtain the best results, specially for AM AE, for which the differences are statistically significant in favour of R 2 .

show abstract

“…2 The most recent approaches combine classification techniques with clustering algorithms to improve the results quality, for example, Hsu uses neural networks applied to brain-computer interface systems, 21 Kodogiannis et al use neural networks and fuzzy clustering for short-term load forecasting 22 and Davis et al combine segmentation and classification for hand radiography. 23 Other research lines have tried to improve these algorithms. For example, some online methods have been developed to avoid the K-means convergence problem to local solutions which depend on the initial values.…”

Section: Clustering Algorithmsmentioning

confidence: 99%

A Genetic Graph-Based Approach for Partitional Clustering

Menéndez

Barrero

Camacho

2014

Int. J. Neur. Syst.

View full text Add to dashboard Cite

Clustering is one of the most versatile tools for data analysis. In the recent years, clustering that seeks the continuity of data (in opposition to classical centroid-based approaches) has attracted an increasing research interest. It is a challenging problem with a remarkable practical interest. The most popular continuity clustering method is the spectral clustering (SC) algorithm, which is based on graph cut: It initially generates a similarity graph using a distance measure and then studies its graph spectrum to find the best cut. This approach is sensitive to the parameters of the metric, and a correct parameter choice is critical to the quality of the cluster. This work proposes a new algorithm, inspired by SC, that reduces the parameter dependency while maintaining the quality of the solution. The new algorithm, named genetic graph-based clustering (GGC), takes an evolutionary approach introducing a genetic algorithm (GA) to cluster the similarity graph. The experimental validation shows that GGC increases robustness of SC and has competitive performance in comparison with classical clustering methods, at least, in the synthetic and real dataset used in the experiments.

show abstract

On the Segmentation and Classification of Hand Radiographs

Cited by 20 publications

References 38 publications

Euler State Networks: Non-dissipative Reservoir Computing

Euler State Networks: Non-dissipative Reservoir Computing

Time series ordinal classification via shapelets

A Genetic Graph-Based Approach for Partitional Clustering

Contact Info

Product

Resources

About