23 results on '"classification ensemble"'
Search Results
2. An ensemble of learned features and reshaping of fractal geometry-based descriptors for classification of histological images.
- Author
-
Roberto, Guilherme Freire, Neves, Leandro Alves, Lumini, Alessandra, Martins, Alessandro Santana, and Nascimento, Marcelo Zanchetta do
- Abstract
Classification of histology images has been the focus of plenty researchers in computer vision. Recently, the most common approaches for this task consist of applying deep learning through CNN models. However, there are some limitations to the use of CNN in the context of histological image classification such as the need for large datasets and the difficulty to implement a generalized model able to classify different types of histology tissue. In this work, an ensemble model based on handcrafted fractal features and deep learning that consists of combining the classification of the ResNet-50 model and the classification of local and global handcrafted features by applying the sum rule is proposed. Fractal geometry concepts are used to obtain handcrafted local and global features from different histological datasets. The local features are reshaped into a matrix in order to compose a feature image. Four different reshaping procedures are evaluated, wherein each generates a representation model of fractal features which is given as input to a CNN model. Another CNN architecture receives as input the original image. After associating the results of both CNN models with the classification of the handcrafted local and global features using machine learning approaches, accuracy rates that range from 88.45% up to 99.77% on five datasets were obtained. Moreover, the model was able to classify images from datasets with different resolutions and imbalanced classes with few training epochs. In general, the proposed method is able to provide results that are compatible with the state-of-the-art in histology image classification. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
3. High-Dimensional Ensemble Learning Classification: An Ensemble Learning Classification Algorithm Based on High-Dimensional Feature Space Reconstruction.
- Author
-
Zhao, Miao and Ye, Ning
- Subjects
MACHINE learning ,CLASSIFICATION algorithms ,FEATURE selection ,NAIVE Bayes classification ,HIGH-dimensional model representation ,CLASSIFICATION ,ALGORITHMS ,PROBLEM solving - Abstract
When performing classification tasks on high-dimensional data, traditional machine learning algorithms often fail to filter out valid information in the features adequately, leading to low levels of classification accuracy. Therefore, this paper explores the high-dimensional data from both the data feature dimension and the model ensemble dimension. We propose a high-dimensional ensemble learning classification algorithm focusing on feature space reconstruction and classifier ensemble, called the HDELC algorithm. First, the algorithm considers feature space reconstruction and then generates a feature space reconstruction matrix. It effectively achieves feature selection and reconstruction for high-dimensional data. An optimal feature space is generated for the subsequent ensemble of the classifier, which enhances the representativeness of the feature space. Second, we recursively determine the number of classifiers and the number of feature subspaces in the ensemble model. Different classifiers in the ensemble system are assigned mutually exclusive non-intersecting feature subspaces for model training. The experimental results show that the HDELC algorithm has advantages compared with most high-dimensional datasets due to its more efficient feature space ensemble capability and relatively reliable ensemble operation performance. The HDELC algorithm makes it possible to solve the classification problem for high-dimensional data effectively and has vital research and application value. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF
4. Boosting EfficientNets Ensemble Performance via Pseudo-Labels and Synthetic Images by pix2pixHD for Infection and Ischaemia Classification in Diabetic Foot Ulcers
- Author
-
Bloch, Louise, Brüngel, Raphael, Friedrich, Christoph M., Goos, Gerhard, Founding Editor, Hartmanis, Juris, Founding Editor, Bertino, Elisa, Editorial Board Member, Gao, Wen, Editorial Board Member, Steffen, Bernhard, Editorial Board Member, Woeginger, Gerhard, Editorial Board Member, Yung, Moti, Editorial Board Member, Yap, Moi Hoon, editor, Cassidy, Bill, editor, and Kendrick, Connah, editor
- Published
- 2022
- Full Text
- View/download PDF
5. High-Dimensional Ensemble Learning Classification: An Ensemble Learning Classification Algorithm Based on High-Dimensional Feature Space Reconstruction
- Author
-
Miao Zhao and Ning Ye
- Subjects
classification ensemble ,feature selection ,high dimensional ,space reconstruction ,ensemble learning ,Technology ,Engineering (General). Civil engineering (General) ,TA1-2040 ,Biology (General) ,QH301-705.5 ,Physics ,QC1-999 ,Chemistry ,QD1-999 - Abstract
When performing classification tasks on high-dimensional data, traditional machine learning algorithms often fail to filter out valid information in the features adequately, leading to low levels of classification accuracy. Therefore, this paper explores the high-dimensional data from both the data feature dimension and the model ensemble dimension. We propose a high-dimensional ensemble learning classification algorithm focusing on feature space reconstruction and classifier ensemble, called the HDELC algorithm. First, the algorithm considers feature space reconstruction and then generates a feature space reconstruction matrix. It effectively achieves feature selection and reconstruction for high-dimensional data. An optimal feature space is generated for the subsequent ensemble of the classifier, which enhances the representativeness of the feature space. Second, we recursively determine the number of classifiers and the number of feature subspaces in the ensemble model. Different classifiers in the ensemble system are assigned mutually exclusive non-intersecting feature subspaces for model training. The experimental results show that the HDELC algorithm has advantages compared with most high-dimensional datasets due to its more efficient feature space ensemble capability and relatively reliable ensemble operation performance. The HDELC algorithm makes it possible to solve the classification problem for high-dimensional data effectively and has vital research and application value.
- Published
- 2024
- Full Text
- View/download PDF
6. Ensemble of RNN Classifiers for Activity Detection Using a Smartphone and Supporting Nodes.
- Author
-
Bernaś, Marcin, Płaczek, Bartłomiej, and Lewandowski, Marcin
- Subjects
- *
SENSOR networks , *WIRELESS sensor networks , *RECURRENT neural networks , *SMARTPHONES , *ATRIOVENTRICULAR node - Abstract
Nowadays, sensor-equipped mobile devices allow us to detect basic daily activities accurately. However, the accuracy of the existing activity recognition methods decreases rapidly if the set of activities is extended and includes training routines, such as squats, jumps, or arm swings. Thus, this paper proposes a model of a personal area network with a smartphone (as a main node) and supporting sensor nodes that deliver additional data to increase activity-recognition accuracy. The introduced personal area sensor network takes advantage of the information from multiple sensor nodes attached to different parts of the human body. In this scheme, nodes process their sensor readings locally with the use of recurrent neural networks (RNNs) to categorize the activities. Then, the main node collects results from supporting sensor nodes and performs a final activity recognition run based on a weighted voting procedure. In order to save energy and extend the network's lifetime, sensor nodes report their local results only for specific types of recognized activity. The presented method was evaluated during experiments with sensor nodes attached to the waist, chest, leg, and arm. The results obtained for a set of eight activities show that the proposed approach achieves higher recognition accuracy when compared with the existing methods. Based on the experimental results, the optimal configuration of the sensor nodes was determined to maximize the activity-recognition accuracy and reduce the number of transmissions from supporting sensor nodes. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
7. High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors.
- Author
-
Wu, Binglong, Shen, Yuan, Guo, Shanxin, Chen, Jinsong, Sun, Luyi, Li, Hongzhong, and Ao, Yong
- Subjects
- *
REMOTE sensing , *DETECTORS , *OBJECT recognition (Computer vision) , *CONVOLUTIONAL neural networks - Abstract
Deep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresolution remote sensing images, where the quality is defined by the Intersection over Union (IoU) threshold used in training. In addition, the imbalance between the positive and negative samples across multiresolution images worsens the detection precision. Recently, it was found that a Cascade region-based convolutional neural network (R-CNN) can potentially achieve a higher quality of detection by introducing a cascaded three-stage structure using progressively improved IoU thresholds. However, the performance of Cascade R-CNN degraded when the fourth stage was added. We investigated the cause and found that the mismatch between the ROI features and the classifier could be responsible for the degradation of performance. Herein, we propose a Cascade R-CNN++ structure to address this issue and extend the three-stage architecture to multiple stages for general use. Specifically, for cascaded classification, we propose a new ensemble strategy for the classifier and region of interest (RoI) features to improve classification accuracy at inference. In localization, we modified the loss function of the bounding box regressor to obtain higher sensitivity around zero. Experiments on the DOTA dataset demonstrated that Cascade R-CNN++ outperforms Cascade R-CNN in terms of precision and detection quality. We conducted further analysis on multiresolution remote sensing images to verify model transferability across different object scales. [ABSTRACT FROM AUTHOR]
- Published
- 2022
- Full Text
- View/download PDF
8. Ensemble of RNN Classifiers for Activity Detection Using a Smartphone and Supporting Nodes
- Author
-
Marcin Bernaś, Bartłomiej Płaczek, and Marcin Lewandowski
- Subjects
mobile phone ,sensor nodes ,activity recognition ,transmission suppression ,recurrent neural network ,classification ensemble ,Chemical technology ,TP1-1185 - Abstract
Nowadays, sensor-equipped mobile devices allow us to detect basic daily activities accurately. However, the accuracy of the existing activity recognition methods decreases rapidly if the set of activities is extended and includes training routines, such as squats, jumps, or arm swings. Thus, this paper proposes a model of a personal area network with a smartphone (as a main node) and supporting sensor nodes that deliver additional data to increase activity-recognition accuracy. The introduced personal area sensor network takes advantage of the information from multiple sensor nodes attached to different parts of the human body. In this scheme, nodes process their sensor readings locally with the use of recurrent neural networks (RNNs) to categorize the activities. Then, the main node collects results from supporting sensor nodes and performs a final activity recognition run based on a weighted voting procedure. In order to save energy and extend the network’s lifetime, sensor nodes report their local results only for specific types of recognized activity. The presented method was evaluated during experiments with sensor nodes attached to the waist, chest, leg, and arm. The results obtained for a set of eight activities show that the proposed approach achieves higher recognition accuracy when compared with the existing methods. Based on the experimental results, the optimal configuration of the sensor nodes was determined to maximize the activity-recognition accuracy and reduce the number of transmissions from supporting sensor nodes.
- Published
- 2022
- Full Text
- View/download PDF
9. A one-class-classification approach to create a stresslevel curve plotter through wearable measurements and behavioral patterns.
- Author
-
Ramírez-Valenzuela, Rodolfo A., Monroy, Raúl, Loyola-González, Octavio, Godínez, Fernando, and Soberanes-Martín, Anabelem
- Abstract
Occupational stress has become an interesting field of research in recent years. Stress in students may yield a decline in academic performance or an increase of a mental issue, hence making of paramount importance the timely diagnosis of stress. Although there exist mechanisms for inferring stress level, most of them: assume the test subject is in a controlled environment; use uncomfortable or unaffordable sensors; or they are applicable only when the subject is at a particular posture. Moreover, to the best of the authors' knowledge, there is no method capable of plotting a person's stress level curve on the fly. In this paper, we propose a method capable of doing so; our method combines a set of one-class-classifiers capable of capturing the user stress level according to four strata (Low, Medium–Low, Medium–High, and High). Throughout our research, we have developed a dataset, called Student Resilience, which contains observation of several test subjects carrying a mobile phone, and wearing a wristband. For each test subject our dataset also contains the output of a collection of tests, especially designed to evaluate mental health and self-perceived stress. We have used the survey output as ground truth for validation purposes. Our method was capable of correctly plotting stress for 87% of the days submitted by the test subjects. Additionally, in a further attempt to validate our method, we have used data mining to determine whether a stress plot is likely to be explained by the unique activities carried out by each test subject for a given day of the week. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
10. High Quality Object Detection for Multiresolution Remote Sensing Imagery Using Cascaded Multi-Stage Detectors
- Author
-
Binglong Wu, Yuan Shen, Shanxin Guo, Jinsong Chen, Luyi Sun, Hongzhong Li, and Yong Ao
- Subjects
object detection ,cascaded detectors ,Intersection over Union (IoU) threshold ,classification ensemble ,bounding box regression ,multiresolution remote sensing images ,Science - Abstract
Deep-learning-based object detectors have substantially improved state-of-the-art object detection in remote sensing images in terms of precision and degree of automation. Nevertheless, the large variation of the object scales makes it difficult to achieve high-quality detection across multiresolution remote sensing images, where the quality is defined by the Intersection over Union (IoU) threshold used in training. In addition, the imbalance between the positive and negative samples across multiresolution images worsens the detection precision. Recently, it was found that a Cascade region-based convolutional neural network (R-CNN) can potentially achieve a higher quality of detection by introducing a cascaded three-stage structure using progressively improved IoU thresholds. However, the performance of Cascade R-CNN degraded when the fourth stage was added. We investigated the cause and found that the mismatch between the ROI features and the classifier could be responsible for the degradation of performance. Herein, we propose a Cascade R-CNN++ structure to address this issue and extend the three-stage architecture to multiple stages for general use. Specifically, for cascaded classification, we propose a new ensemble strategy for the classifier and region of interest (RoI) features to improve classification accuracy at inference. In localization, we modified the loss function of the bounding box regressor to obtain higher sensitivity around zero. Experiments on the DOTA dataset demonstrated that Cascade R-CNN++ outperforms Cascade R-CNN in terms of precision and detection quality. We conducted further analysis on multiresolution remote sensing images to verify model transferability across different object scales.
- Published
- 2022
- Full Text
- View/download PDF
11. Machine Learning Classification Ensemble of Multitemporal Sentinel-2 Images: The Case of a Mixed Mediterranean Ecosystem
- Author
-
Christos Vasilakos, Dimitris Kavroudakis, and Aikaterini Georganta
- Subjects
remote sensing ,classification ensemble ,machine learning ,Sentinel-2 ,geographic information system (GIS) ,Science - Abstract
Land cover type classification still remains an active research topic while new sensors and methods become available. Applications such as environmental monitoring, natural resource management, and change detection require more accurate, detailed, and constantly updated land-cover type mapping. These needs are fulfilled by newer sensors with high spatial and spectral resolution along with modern data processing algorithms. Sentinel-2 sensor provides data with high spatial, spectral, and temporal resolution for the in classification of highly fragmented landscape. This study applies six traditional data classifiers and nine ensemble methods on multitemporal Sentinel-2 image datasets for identifying land cover types in the heterogeneous Mediterranean landscape of Lesvos Island, Greece. Support vector machine, random forest, artificial neural network, decision tree, linear discriminant analysis, and k-nearest neighbor classifiers are applied and compared with nine ensemble classifiers on the basis of different voting methods. kappa statistic, F1-score, and Matthews correlation coefficient metrics were used in the assembly of the voting methods. Support vector machine outperformed the base classifiers with kappa of 0.91. Support vector machine also outperformed the ensemble classifiers in an unseen dataset. Five voting methods performed better than the rest of the classifiers. A diversity study based on four different metrics revealed that an ensemble can be avoided if a base classifier shows an identifiable superiority. Therefore, ensemble approaches should include a careful selection of base-classifiers based on a diversity analysis.
- Published
- 2020
- Full Text
- View/download PDF
12. Ensemble machine learning algorithm optimization of bankruptcy prediction of bank
- Author
-
Bambang Siswoyo, Zuraida Abal Abas, Ahmad Naim Che Pee, Rita Komalasari, and Nano Suryana
- Subjects
Bankruptcy ,Information Systems and Management ,Class imbalance ,Classification Ensemble ,Artificial Intelligence ,Control and Systems Engineering ,Bagging ,Electrical and Electronic Engineering - Abstract
The ensemble consists of a single set of individually trained models, the predictions of which are combined when classifying new cases, in building a good classification model requires the diversity of a single model. The algorithm, logistic regression, support vector machine, random forest, and neural network are single models as alternative sources of diversity information. Previous research has shown that ensembles are more accurate than single models. Single model and modified ensemble bagging model are some of the techniques we will study in this paper. We experimented with the banking industry’s financial ratios. The results of his observations are: First, an ensemble is always more accurate than a single model. Second, we observe that modified ensemble bagging models show improved classification model performance on balanced datasets, as they can adjust behavior and make them more suitable for relatively small datasets. The accuracy rate is 97% in the bagging ensemble learning model, an increase in the accuracy level of up to 16% compared to other models that use unbalanced datasets.
- Published
- 2022
13. Photogrammetric UAV Mapping of Terrain under Dense Coastal Vegetation: An Object-Oriented Classification Ensemble Algorithm for Classification and Terrain Correction.
- Author
-
Xuelian Meng, Nan Shang, Xukai Zhang, Chunyan Li, Kaiguang Zhao, Xiaomin Qiu, and Weeks, Eddie
- Subjects
- *
TERRAIN mapping , *PHOTOGRAMMETRY , *COASTAL plants , *DRONE photography , *GROUND vegetation cover , *MACHINE learning - Abstract
Photogrammetric UAV sees a surge in use for high-resolution mapping, but its use to map terrain under dense vegetation cover remains challenging due to a lack of exposed ground surfaces. This paper presents a novel object-oriented classification ensemble algorithm to leverage height, texture and contextual information of UAV data to improve landscape classification and terrain estimation. Its implementation incorporates multiple heuristics, such as multi-input machine learning-based classification, object-oriented ensemble, and integration of UAV and GPS surveys for terrain correction. Experiments based on a densely vegetated wetland restoration site showed classification improvement from 83.98% to 96.12% in overall accuracy and from 0.7806 to 0.947 in kappa value. Use of standard and existing UAV terrain mapping algorithms and software produced reliable digital terrain model only over exposed bare grounds (mean error = -0.019 m and RMSE = 0.035 m) but severely overestimated the terrain by ~80% of mean vegetation height in vegetated areas. The terrain correction method successfully reduced the mean error from 0.302 m to -0.002 m (RMSE from 0.342 m to 0.177 m) in low vegetation and from 1.305 m to 0.057 m (RMSE from 1.399 m to 0.550 m) in tall vegetation. Overall, this research validated a feasible solution to integrate UAV and RTK GPS for terrain mapping in densely vegetated environments. [ABSTRACT FROM AUTHOR]
- Published
- 2017
- Full Text
- View/download PDF
14. Spectral–Spatial Rotation Forest for Hyperspectral Image Classification.
- Author
-
Xia, Junshi, Bombrun, Lionel, Berthoumieu, Yannick, Germain, Christian, and Du, Peijun
- Abstract
Rotation Forest (RoF) is a recent powerful decision tree (DT) ensemble classifier of hyperspectral images. RoF exploits random feature selection and data transformation techniques to improve both the diversity and accuracy of DT classifiers. Conventional RoF only considers data transformation on spectral information. To overcome this limitation, we propose a spectral and spatial RoF (SSRoF), to further improve the performance. In SSRoF, pixels are first smoothed by the multiscale (MS) spatial weight mean filtering. Then, spectral–spatial data transformation, which is based on a joint spectral and spatial rotation matrix, is introduced into the RoF. Finally, classification results obtained from each scale are integrated by a majority voting rule. Experimental results on two datasets indicate the competitive performance of the proposed method when compared to other state-of-the-art methods. [ABSTRACT FROM PUBLISHER]
- Published
- 2017
- Full Text
- View/download PDF
15. Ensemble of handcrafted and deep learning features based on fractal geometry for the classification of histology images
- Author
-
Guilherme Freire Roberto, Neves, Leandro Alves, Nascimento, Marcelo Zanchetta do, Barioni, Maria Camila Nardini, Julia, Rita Maria da Silva, Lorena, Ana Carolina, and Ramos, Rodrigo Pereira
- Subjects
aprendizado profundo ,comitê de classificadores ,histology images ,Imagens - Classificação ,CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO [CNPQ] ,deep learning ,imagens histológicas ,Aprendizado do computador ,Computação ,atributos fractais ,classification ensemble ,Análise de imagem - Processamento eletrônico de dados ,fractal features - Abstract
CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior CNPq - Conselho Nacional de Desenvolvimento Científico e Tecnológico FAPEMIG - Fundação de Amparo a Pesquisa do Estado de Minas Gerais A classificação de imagens histológicas é uma tarefa que tem sido amplamente explorada nas recentes pesquisas de visão computacional. A abordagem mais estudada para esta tarefa tem sido a aplicação de aprendizado profundo por meio de modelos de CNN. Entretanto, o uso de CNN no contexto da classificação de imagens histológicas tem ainda algumas limitações, como a necessidade de grandes conjuntos de dados e a dificuldade de implementar um modelo generalizado capaz de classificar diferentes tipos de tecido histológico. Neste trabalho, propõe-se um modelo de comitê de classificadores baseado em atributos fractais e aprendizado profundo que consiste em combinar a classificação de duas CNN e a classificação de atributos manuais locais e globais aplicando a regra da soma. A extração das características é aplicada para obter 300 atributos fractais de diferentes conjuntos de dados histológicos. Estes atributos são reorganizados em uma matriz a fim de compor uma imagem RGB. São avaliados quatro procedimentos diferentes para efetuar esta reorganização, que geram modelos de representação dos atributos fractais que são dados como entrada para uma primeira CNN. Outra CNN recebe como entrada a imagem original correspondente. Depois de combinar os resultados de ambas as CNN com a classificação dos atributos manuais utilizando abordagens clássicas de aprendizado de máquina, foram obtidas acurácias que variam de 88,45\% a 99,77\% em cinco conjuntos de dados diferentes. Além disso, o modelo foi capaz de classificar imagens de conjuntos de dados com classes desbalanceadas, sem a necessidade de imagens possuírem a mesma resolução, e com um treinamento de 10 épocas. Também foi verificado que os resultados obtidos são complementares aos estudos mais relevantes publicados recentemente no contexto da classificação de imagens histológicas. Classification of histology images is a task that has been widely explored on recent computer vision researches. The most studied approach for this task has been the application of deep learning through CNN models. However, the use of CNN in the context of histological images classification has yet some limitations such as the need of large datasets and the difficulty to implement a generalized model able to classify different types of histology tissue. In this project, an ensemble model based on handcrafted fractal features and deep learning that consists on combining the classification of two CNN and the classification of local and global handcrafted features by applying the sum rule is proposed. Feature extraction is applied to obtain 300 fractal features from different histological datasets. These features are reshaped into a matrix in order to compose an RGB feature image. Four different reshaping procedures are evaluated, wherein each generates a representation model of fractal features which is given as input to a CNN. Another CNN receives as input the correspondent original image. After combining the results of both CNN with the classification of the handcrafted features using classical machine learing approaches, accuracies that range from 88.45\% up to 99.77\% on five different datasets were obtained. Moreover, the model was able to classify images from datasets with imbalanced classes, without the need of images having the same resolution, and using 10 epochs for training. It was also verified that the obtained results are complementary to the most relevant studies recently published in the context of histology image classification. Tese (Doutorado) 2023-12-08
- Published
- 2021
16. Fractal Neural Network: A new ensemble of fractal geometry and convolutional neural networks for the classification of histology images
- Author
-
Marcelo Zanchetta do Nascimento, Alessandra Lumini, Leandro Alves Neves, Guilherme Freire Roberto, Universidade Federal de Uberlândia (UFU), FC, Universidade Estadual Paulista (Unesp), Roberto G.F., Lumini A., Neves L.A., and do Nascimento M.Z.
- Subjects
0209 industrial biotechnology ,Computer science ,Classification ensemble ,Feature extraction ,Context (language use) ,02 engineering and technology ,Convolutional neural network ,020901 industrial engineering & automation ,Fractal ,Artificial Intelligence ,0202 electrical engineering, electronic engineering, information engineering ,Fractal feature ,Contextual image classification ,Artificial neural network ,Ensemble forecasting ,business.industry ,Deep learning ,General Engineering ,Pattern recognition ,Computer Science Applications ,Histology images ,Fractal features ,020201 artificial intelligence & image processing ,Artificial intelligence ,business - Abstract
Made available in DSpace on 2021-06-25T10:35:39Z (GMT). No. of bitstreams: 0 Previous issue date: 2021-03-15 Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) Classification of histology images is a task that has been widely explored on recent computer vision researches. The most studied approach for this task has been the application of deep learning through a convolutional neural network (CNN) model. However, the use of CNNs in the context of histological images classification has yet some limitations such as the need of large datasets, the slow training time and the difficult to implement a generalized model able to classify different types of histology tissues. In this paper, we propose an ensemble model based on handcrafted fractal features and deep learning that consists of combining the classification of two CNNs by applying the sum rule. We apply feature extraction to obtain 300 fractal features from different histological datasets. These features are reshaped into a 10×10×3 matrix to compose an artificial image that is given as input to the first CNN. The second CNN model receives as input the correspondent original image. After combining the results of both CNNs, accuracies that range from 89.66% up to 99.62% were obtained from five different datasets. Moreover, our model was able to classify images from datasets with imbalanced classes, without the need for images having the same resolution, and in relative fast training time. We also verified that the obtained results are compatible with the most recent and relevant studies recently published in the context of histology image classification. Faculty of Computer Science (FACOM) - Federal University of Uberlândia (UFU) Av. João Naves de Ávila 2121 BLB 38400-902 Uberlândia MG Department of Computer Science and Engineering (DISI) - University of Bologna Via dell'Università 50 47521 Cesena FC Department of Computer Science and Statistics (DCCE) São Paulo State University (UNESP) R. Cristóvão Colombo 2265 15054-000 São José do Rio Preto SP Department of Computer Science and Statistics (DCCE) São Paulo State University (UNESP) R. Cristóvão Colombo 2265 15054-000 São José do Rio Preto SP CNPq: #304848/2018-2 CNPq: #313365/2018-0 CNPq: #430965/2018-4 CAPES: #88882.429128/2019-01 FAPEMIG: #APQ-00578-18
- Published
- 2021
17. An Intelligent Anti-phishing Strategy Model for Phishing Website Detection.
- Author
-
Zhuang, Weiwei, Jiang, Qingshan, and Xiong, Tengke
- Abstract
As a new form of malicious software, phishing websites appear frequently in recent years, which cause great harm to online financial services and data security. In this paper, we design and implement an intelligent model for detecting phishing websites. In this model, we extract 10 different types of features such as title, keyword and link text information to represent the website. Heterogeneous classifiers are then built based on these different features. We propose a principled ensemble classification algorithm to combine the predicted results from different phishing detection classifiers. Hierarchical clustering technique has been employed for automatic phishing categorization. Case studies on large and real daily phishing websites collected from King soft Internet Security Lab demonstrate that our proposed model outperforms other commonly used anti-phishing methods and tools in phishing website detection. [ABSTRACT FROM PUBLISHER]
- Published
- 2012
- Full Text
- View/download PDF
18. Automatic post-picking using MAPPOS improves particle image detection from cryo-EM micrographs.
- Author
-
Norousi, Ramin, Wickles, Stephan, Leidig, Christoph, Becker, Thomas, Schmid, Volker J., Beckmann, Roland, and Tresch, Achim
- Subjects
- *
ELECTRON microscopy , *CRYOBIOCHEMISTRY , *MACROMOLECULAR dynamics , *MICROGRAPHICS , *COMPARATIVE studies , *PARTICLE image velocimetry - Abstract
Abstract: Cryo-electron microscopy (cryo-EM) studies using single particle reconstruction are extensively used to reveal structural information on macromolecular complexes. Aiming at the highest achievable resolution, state of the art electron microscopes automatically acquire thousands of high-quality micrographs. Particles are detected on and boxed out from each micrograph using fully- or semi-automated approaches. However, the obtained particles still require laborious manual post-picking classification, which is one major bottleneck for single particle analysis of large datasets. We introduce MAPPOS, a supervised post-picking strategy for the classification of boxed particle images, as additional strategy adding to the already efficient automated particle picking routines. MAPPOS employs machine learning techniques to train a robust classifier from a small number of characteristic image features. In order to accurately quantify the performance of MAPPOS we used simulated particle and non-particle images. In addition, we verified our method by applying it to an experimental cryo-EM dataset and comparing the results to the manual classification of the same dataset. Comparisons between MAPPOS and manual post-picking classification by several human experts demonstrated that merely a few hundred sample images are sufficient for MAPPOS to classify an entire dataset with a human-like performance. MAPPOS was shown to greatly accelerate the throughput of large datasets by reducing the manual workload by orders of magnitude while maintaining a reliable identification of non-particle images. [Copyright &y& Elsevier]
- Published
- 2013
- Full Text
- View/download PDF
19. Combining multiple predictive models using genetic algorithms.
- Author
-
Janusz, Andrzej
- Subjects
- *
GENETIC algorithms , *PREDICTION models , *MATHEMATICAL models , *REGRESSION analysis , *DATA mining - Abstract
Blending is a well-established technique, commonly used to increase performance of predictive models. Its effectiveness has been confirmed in practice as most of the latest international data-mining contest winners were using some kind of a committee of classifiers to produce their final entry. This paper presents a method of using a genetic algorithm to optimize an ensemble of multiple classification or regression models. An implementation of that method in R system, called Genetic Meta-Blender, was tested during the Australasian Data Mining 2009 Analytic Challenge. A subject of this data mining competition was the methods for combining predictive models. The described approach was awarded with the Grand Champion prize for achieving the best overall result. In this paper, the purpose of the challenge is described and details of the winning approach are given. The results of Genetic Meta-Blender are also discussed and compared to several baseline scores. Additionally, GMB is evaluated on data from a different data mining competition, namely SIAM SDM'11 Contest: Prediction of Biological Properties of Molecules from Chemical Structure. [ABSTRACT FROM AUTHOR]
- Published
- 2012
- Full Text
- View/download PDF
20. Fractal Neural Network: A new ensemble of fractal geometry and convolutional neural networks for the classification of histology images.
- Author
-
Roberto, Guilherme Freire, Lumini, Alessandra, Neves, Leandro Alves, and do Nascimento, Marcelo Zanchetta
- Subjects
- *
CONVOLUTIONAL neural networks , *FRACTALS , *FRACTAL analysis , *CLASSIFICATION , *COMPUTER vision , *HISTOLOGY - Abstract
Classification of histology images is a task that has been widely explored on recent computer vision researches. The most studied approach for this task has been the application of deep learning through a convolutional neural network (CNN) model. However, the use of CNNs in the context of histological images classification has yet some limitations such as the need of large datasets, the slow training time and the difficult to implement a generalized model able to classify different types of histology tissues. In this paper, we propose an ensemble model based on handcrafted fractal features and deep learning that consists of combining the classification of two CNNs by applying the sum rule. We apply feature extraction to obtain 300 fractal features from different histological datasets. These features are reshaped into a 10 × 10 × 3 matrix to compose an artificial image that is given as input to the first CNN. The second CNN model receives as input the correspondent original image. After combining the results of both CNNs, accuracies that range from 89.66% up to 99.62% were obtained from five different datasets. Moreover, our model was able to classify images from datasets with imbalanced classes, without the need for images having the same resolution, and in relative fast training time. We also verified that the obtained results are compatible with the most recent and relevant studies recently published in the context of histology image classification. • A CNN ensemble that requires few training epochs for classifying histology images. • Fractal dimension, lacunarity and percolation features are extracted from the images. • Fractal features are reshaped as a synthetic image and then given as input to a CNN. • Accuracies up to 99.62% were obtained from classifying 5 histology image datasets. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF
21. Machine Learning Classification Ensemble of Multitemporal Sentinel-2 Images: The Case of a Mixed Mediterranean Ecosystem.
- Author
-
Vasilakos, Christos, Kavroudakis, Dimitris, and Georganta, Aikaterini
- Subjects
- *
FISHER discriminant analysis , *NATURAL resources management , *SUPPORT vector machines , *ARTIFICIAL neural networks , *FRAGMENTED landscapes , *ELECTRONIC voting , *MACHINE learning , *K-nearest neighbor classification - Abstract
Land cover type classification still remains an active research topic while new sensors and methods become available. Applications such as environmental monitoring, natural resource management, and change detection require more accurate, detailed, and constantly updated land-cover type mapping. These needs are fulfilled by newer sensors with high spatial and spectral resolution along with modern data processing algorithms. Sentinel-2 sensor provides data with high spatial, spectral, and temporal resolution for the in classification of highly fragmented landscape. This study applies six traditional data classifiers and nine ensemble methods on multitemporal Sentinel-2 image datasets for identifying land cover types in the heterogeneous Mediterranean landscape of Lesvos Island, Greece. Support vector machine, random forest, artificial neural network, decision tree, linear discriminant analysis, and k-nearest neighbor classifiers are applied and compared with nine ensemble classifiers on the basis of different voting methods. kappa statistic, F1-score, and Matthews correlation coefficient metrics were used in the assembly of the voting methods. Support vector machine outperformed the base classifiers with kappa of 0.91. Support vector machine also outperformed the ensemble classifiers in an unseen dataset. Five voting methods performed better than the rest of the classifiers. A diversity study based on four different metrics revealed that an ensemble can be avoided if a base classifier shows an identifiable superiority. Therefore, ensemble approaches should include a careful selection of base-classifiers based on a diversity analysis. [ABSTRACT FROM AUTHOR]
- Published
- 2020
- Full Text
- View/download PDF
22. Photogrammetric UAV Mapping of Terrain under Dense Coastal Vegetation: An Object-Oriented Classification Ensemble Algorithm for Classification and Terrain Correction
- Author
-
Xukai Zhang, Kaiguang Zhao, Eddie Weeks, Chunyan Li, Xuelian Meng, Nan Shang, and Xiaomin Qiu
- Subjects
010504 meteorology & atmospheric sciences ,Mean squared error ,Science ,0211 other engineering and technologies ,coastal topographic mapping ,Terrain ,02 engineering and technology ,01 natural sciences ,wetland restoration ,classification correction ,Software ,Real Time Kinematic ,Leverage (statistics) ,Digital elevation model ,021101 geological & geomatics engineering ,0105 earth and related environmental sciences ,Remote sensing ,photogrammetric UAV ,high resolution ,terrain correction ,object-oriented analysis ,classification ensemble ,business.industry ,Photogrammetry ,Global Positioning System ,General Earth and Planetary Sciences ,business ,Algorithm ,Geology - Abstract
Photogrammetric UAV sees a surge in use for high-resolution mapping, but its use to map terrain under dense vegetation cover remains challenging due to a lack of exposed ground surfaces. This paper presents a novel object-oriented classification ensemble algorithm to leverage height, texture and contextual information of UAV data to improve landscape classification and terrain estimation. Its implementation incorporates multiple heuristics, such as multi-input machine learning-based classification, object-oriented ensemble, and integration of UAV and GPS surveys for terrain correction. Experiments based on a densely vegetated wetland restoration site showed classification improvement from 83.98% to 96.12% in overall accuracy and from 0.7806 to 0.947 in kappa value. Use of standard and existing UAV terrain mapping algorithms and software produced reliable digital terrain model only over exposed bare grounds (mean error = −0.019 m and RMSE = 0.035 m) but severely overestimated the terrain by ~80% of mean vegetation height in vegetated areas. The terrain correction method successfully reduced the mean error from 0.302 m to −0.002 m (RMSE from 0.342 m to 0.177 m) in low vegetation and from 1.305 m to 0.057 m (RMSE from 1.399 m to 0.550 m) in tall vegetation. Overall, this research validated a feasible solution to integrate UAV and RTK GPS for terrain mapping in densely vegetated environments.
- Published
- 2017
- Full Text
- View/download PDF
23. Combined pattern search optimization of feature extraction and classification parameters in facial recognition
- Author
-
Sorin Moga, Xia Mao, Ctlin-Daniel Cleanu, Gilbert Pradel, Yuli Xue, Faculty of Electronics and Telecommunications, University 'POLITEHNICA' Timişoara, School of Electronic and Information Engineering, Beihang University, HANDS, Informatique, Biologie Intégrative et Systèmes Complexes ( IBISC ), Université d'Évry-Val-d'Essonne ( UEVE ) -Université d'Évry-Val-d'Essonne ( UEVE ), Lab-STICC_TB_CID_DECIDE, Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance ( Lab-STICC ), École Nationale d'Ingénieurs de Brest ( ENIB ) -Université de Bretagne Sud ( UBS ) -Université de Brest ( UBO ) -Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques ( IBNM ), Université de Brest ( UBO ) -Université européenne de Bretagne ( UEB ) -ENSTA Bretagne-Institut Mines-Télécom [Paris]-Centre National de la Recherche Scientifique ( CNRS ) -École Nationale d'Ingénieurs de Brest ( ENIB ) -Université de Bretagne Sud ( UBS ) -Université de Brest ( UBO ) -Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques ( IBNM ), Université de Brest ( UBO ) -Université européenne de Bretagne ( UEB ) -ENSTA Bretagne-Institut Mines-Télécom [Paris]-Centre National de la Recherche Scientifique ( CNRS ), Département Logique des Usages, Sciences sociales et Sciences de l'Information ( LUSSI ), Université européenne de Bretagne ( UEB ) -Télécom Bretagne-Institut Mines-Télécom [Paris], Beihang University (BUAA), Informatique, Biologie Intégrative et Systèmes Complexes (IBISC), Université d'Évry-Val-d'Essonne (UEVE), Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance (Lab-STICC), Université européenne de Bretagne - European University of Brittany (UEB)-École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques (IBNM), Université de Brest (UBO)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS)-Université européenne de Bretagne - European University of Brittany (UEB)-École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques (IBNM), Université de Brest (UBO)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS), Département Logique des Usages, Sciences sociales et Sciences de l'Information (LUSSI), Université européenne de Bretagne - European University of Brittany (UEB)-Télécom Bretagne-Institut Mines-Télécom [Paris] (IMT), École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques (IBNM), Université de Brest (UBO)-Université européenne de Bretagne - European University of Brittany (UEB)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS)-École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Télécom Bretagne-Institut Brestois du Numérique et des Mathématiques (IBNM), Université de Brest (UBO)-Université européenne de Bretagne - European University of Brittany (UEB)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS), and Institut Mines-Télécom [Paris] (IMT)-Télécom Bretagne-Université européenne de Bretagne - European University of Brittany (UEB)
- Subjects
[ INFO.INFO-TS ] Computer Science [cs]/Signal and Image Processing ,Computer science ,Feature extraction ,0211 other engineering and technologies ,Linear classifier ,02 engineering and technology ,computer.software_genre ,Facial recognition system ,Pattern search ,k-nearest neighbors algorithm ,[INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing ,Artificial Intelligence ,0202 electrical engineering, electronic engineering, information engineering ,Feature (machine learning) ,Face recognition ,Sparse representation ,021103 operations research ,business.industry ,Kanade–Lucas–Tomasi feature tracker ,Pattern recognition ,Sparse approximation ,classification ensemble ,ComputingMethodologies_PATTERNRECOGNITION ,Feature (computer vision) ,Signal Processing ,020201 artificial intelligence & image processing ,Computer Vision and Pattern Recognition ,Artificial intelligence ,Data mining ,business ,Pattern search optimization ,computer ,Classifier (UML) ,Software - Abstract
International audience; Constantly, the assumption is made that there is an independent contribution of the individual feature extraction and classifier parameters to the recognition performance. In our approach, the problems of feature extraction and classifier design are viewed together as a single matter of estimating the optimal parameters from limited data. We propose, for the problem of facial recognition, a combination between an Interest Operator based feature extraction technique and a k-NN statistical classifier having the parameters determined using a pattern search based optimization technique. This approach enables us to achieve both higher classification accuracy and faster processing time.
- Published
- 2011
- Full Text
- View/download PDF
Catalog
Discovery Service for Jio Institute Digital Library
For full access to our library's resources, please sign in.