Efficient Classification Approach Research Articles

Au sujet de l’article : Petitdant B. Origines, histoire, évolutions de la mesure de la force de préhension et des dynamomètres médicaux. Kinesither Rev 2017;17(181):40–58

In many areas, the volume of text information is increasing rapidly, thereby demanding efficient text classification approaches. Several methods are available at present, but most exhibit declining performance as the dimensionality of the problem increases, or they incur high computational costs for training, which limit their application in real scenarios. Thus, it is necessary to develop a method that can process high dimensional data in a rapid manner. In this study, we propose the MDLText, an efficient, lightweight, scalable, and fast multinomial text classifier, which is based on the minimum description length principle. MDLText exhibits fast incremental learning as well as being sufficiently robust to prevent overfitting, which are desirable features in real-world applications, large-scale problems, and online scenarios. Our experiments were carefully designed to ensure that we obtained statistically sound results, which demonstrated that the proposed approach achieves a good balance between predictive power and computational efficiency.

Improved Audio Scene Classification Based on Label-Tree Embeddings and Convolutional Neural Networks

In this paper, we present an efficient approach for audio scene classification. We aim at learning representations for scene examples by exploring the structure of their class labels. A category taxonomy is automatically learned by collectively optimizing a tree-structured clustering of the given labels into multiple metaclasses. A scene recording is then transformed into a label-tree embedding image. Elements of the image represent the likelihoods that the scene instance belongs to the metaclasses. We investigate classification with label-tree embedding features learned from different low-level features as well as their fusion. We show that the combination of multiple features is essential to obtain good performance. While averaging label-tree embedding images over time yields good performance, we argue that average pooling possesses an intrinsic shortcoming. We alternatively propose an improved classification scheme to bypass this limitation. We aim at automatically learning common templates that are useful for the classification task from these images using simple but tailored convolutional neural networks. The trained networks are then employed as a feature extractor that matches the learned templates across a label-tree embedding image and produce the maximum matching scores as features for classification. Since audio scenes exhibit rich content, template learning and matching on low-level features would be inefficient. With label-tree embedding features, we have quantized and reduced the low-level features into the likelihoods of the metaclasses, on which the template learning and matching are efficient. We study both training convolutional neural networks on stacked label-tree embedding images and multistream networks. Experimental results on the DCASE2016 and LITIS Rouen datasets demonstrate the efficiency of the proposed methods.

Object- and pixel-based classifications of macroalgae farming area with high spatial resolution imagery

Macroalgae plays an important role in coastal ecosystems. The accurate delineation of macroalgae areas is important for environmental management. This study compared the pixel- and object-based methods using Gaofen satellite no. 2 image to explore an efficient classification approach. Expert system rules and nearest neighbour classifier were adopted for object-based classification, whereas maximum likelihood classifier was implemented in the pixel-based approach. Normalized difference vegetation index, normalized difference water index, mean value of the blue band and geometric characteristics were selected as features to distinguish macroalgae farms by considering the spectral and spatial characteristics. Results show that the object-based method achieved a higher overall accuracy and kappa coefficient than the pixel-based method. Moreover, the object-based approach displayed superiority in identifying Porphyra class. These findings suggest that the object-based method can delineate macroalgae farming areas efficiently and be applied in the future to monitor the macroalgae farms with high spatial resolution imagery.

Classification of multiple diseases based on wavelet features

This study presents an efficient disease classification approach based on medical images. The approach is more efficient as it reduces the computational complexity. The implementation uses only two wavelet filters in selecting the texture features as compared with five filters used in the earlier research works. The computed average and energy features are fed to feed-forward neural network (FFNN) and support vector machine (SVM) classifiers. The SVM is proved as a better classifier than the FFNN for all the three diseases related to skin, breast and retina with an improved accuracies of 89%, 92% and 100%, respectively. Also, a graphical user interface is developed useful for various disease classification based on the whole dataset of size 100.

An edge-based method for effective abandoned luggage detection in complex surveillance videos

Abandoned objects detection is one of the most challenging tasks in intelligent video surveillance systems. In this paper we present a new method for detecting abandoned objects (AO) using edges instead of pixel intensities. Our main focus, is on reducing false alarms, while keeping a high positives detection rate. Based on edge information, the proposed method reduces errors rate considerably compared to pixel intensities based approaches. At first, static edges are detected by applying a temporal accumulation step using the foreground edges mask resulting from the edge-based background subtraction model. Then, edges clustering is applied on the obtained stable edges mask, using edges’ position and stability in time to delineate the object bounding box. Finally, an efficient classification approach, relying on edges’ position, orientation, and staticness based scores, is applied on the AO candidates. The proposed approach has been validated on several challenging benchmarks, and is compared to other works of the literature. The results shows that our method reduce false alarms rates greatly, while keeping a good detection accuracy of true positives.

MDLText: An efficient and lightweight text classifier

In many areas, the volume of text information is increasing rapidly, thereby demanding efficient text classification approaches. Several methods are available at present, but most exhibit declining performance as the dimensionality of the problem increases, or they incur high computational costs for training, which limit their application in real scenarios. Thus, it is necessary to develop a method that can process high dimensional data in a rapid manner. In this study, we propose the MDLText, an efficient, lightweight, scalable, and fast multinomial text classifier, which is based on the minimum description length principle. MDLText exhibits fast incremental learning as well as being sufficiently robust to prevent overfitting, which are desirable features in real-world applications, large-scale problems, and online scenarios. Our experiments were carefully designed to ensure that we obtained statistically sound results, which demonstrated that the proposed approach achieves a good balance between predictive power and computational efficiency.

Privacy-preserved big data analysis based on asymmetric imputation kernels and multiside similarities

This study presents an efficient approach for incomplete data classification, where the entries of samples are missing or masked due to privacy preservation. To deal with these incomplete data, a new kernel function with asymmetric intrinsic mappings is proposed in this study. Such a new kernel uses three-side similarities for kernel matrix formation. The similarity between a testing instance and a training sample relies not only on their distance but also on the relation between the testing sample and the centroid of the class, where the training sample belongs. This reduces biased estimation compared with typical methods when only one training sample is used for kernel matrix formation. Furthermore, centroid generation does not involve any clustering algorithms. The proposed kernel is capable of performing data imputation by using class-dependent averages. This enhances Fisher Discriminant Ratios and data discriminability. Experiments on two open databases were carried out for evaluating the proposed method. The result indicated that the accuracy of the proposed method was higher than that of the baseline. These findings thereby demonstrated the effectiveness of the proposed idea.

Image Classification Using Fusion of Holistic Visual Descriptions

An efficient approach for scene classification is necessary for automatically labeling an image as well as for retrieval of desired images from large scale repositories. In this paper machine learning and computer vision techniques have been applied for scene classification. The system is based on feature fusion method with holistic visual color, texture and edge descriptors. Color moments, Color Coherence Vector, Color Auto Correlogram, GLCM, Daubechies Wavelets, Gabor filters and MPEG-7 Edge Direction Histogram have been used in the proposed system to find the best combination of features for this problem. Two state-of- the-art soft computing machine learning techniques: Support vector machine (SVM) and Artificial Neural Networks have been used to classify scene images into meaningful categories. The benchmarked Oliva-Torralba dataset has been used in this research. We report satisfactory categorization performances on a large data set of eight categories of 2688 complex, natural and urban scenes. Using a set of exhaustive experiments our proposed system has achieved classification accuracy as high as 92.5% for natural scenes (OT4) and as high as 86.4% for mixed scene categories (OT8). We also evaluate the system performance by predictive accuracy measures namely sensitivity, specificity, F-score and kappa statistic.

An Efficient Approach for Segmentation, Feature Extraction and Classification of Audio Signals

Due to the presence of non-stationarities and discontinuities in the audio signal, segmentation and classification of audio signal is a really challenging task. Automatic music classification and annotation is still considered as a challenging task due to the difficulty of extracting and selecting the optimal audio features. Hence, this paper proposes an efficient approach for segmentation, feature extraction and classification of audio signals. Enhanced Mel Frequency Cepstral Coefficient (EMFCC)-Enhanced Power Normalized Cepstral Coefficients (EPNCC) based feature extraction is applied for the extraction of features from the audio signal. Then, multi-level classification is done to classify the audio signal as a musical or non-musical signal. The proposed approach achieves better performance in terms of precision, Normalized Mutual Information (NMI), F-score and entropy. The PNN classifier shows high False Rejection Rate (FRR), False Acceptance Rate (FAR), Genuine Acceptance rate (GAR), sensitivity, specificity and accuracy with respect to the number of classes.

Hybridizing firefly algorithms with a probabilistic neural network for solving classification problems

Classification is one of the important tasks in data mining. The probabilistic neural network (PNN) is a well-known and efficient approach for classification. The objective of the work presented in this paper is to build on this approach to develop an effective method for classification problems that can find high-quality solutions (with respect to classification accuracy) at a high convergence speed. To achieve this objective, we propose a method that hybridizes the firefly algorithm with simulated annealing (denoted as SFA), where simulated annealing is applied to control the randomness step inside the firefly algorithm while optimizing the weights of the standard PNN model. We also extend our work by investigating the effectiveness of using Lévy flight within the firefly algorithm (denoted as LFA) to better explore the search space and by integrating SFA with Lévy flight (denoted as LSFA) in order to improve the performance of the PNN. The algorithms were tested on 11 standard benchmark datasets. Experimental results indicate that the LSFA shows better performance than the SFA and LFA. Moreover, when compared with other algorithms in the literature, the LSFA is able to obtain better results in terms of classification accuracy.

Long-term epileptic EEG classification via 2D mapping and textural features

Interpretation of long-term Electroencephalography (EEG) records is a tiresome task for clinicians. This paper presents an efficient, low cost and novel approach for patient-specific classification of long-term epileptic EEG records. We aim to achieve this with the minimum supervision from the neurologist. To accomplish this objective, first a novel feature extraction method is proposed based on the mapping of EEG signals into two dimensional space, resulting into a texture image. The texture image is constructed by mapping and scaling EEG signals and their associated frequency sub-bands into the gray-level image domain. Image texture analysis using gray level co-occurrence matrix (GLCM) is then applied in order to extract multivariate features which are able to differentiate between seizure and seizure-free events. To evaluate the discriminative power of the proposed feature extraction method, a comparative study is performed, against other dedicated feature extraction methods. The comparative performance evaluations show that the proposed feature extraction method can outperform other state-of-art feature extraction methods with a low computational cost. With a training rate of 25%, the overall sensitivity of 70.19% and specificity of 97.74% are achieved in the classification of over 163h of EEG records using support vector machine (SVM) classifiers with linear kernels and trained by the stochastic gradient descent (SGD) algorithm.

Classification of mammogram using two-dimensional discrete orthonormal S-transform for breast cancer detection.

An efficient approach for classification of mammograms for detection of breast cancer is presented. The approach utilises the two-dimensional discrete orthonormal S-transform (DOST) to extract the coefficients from the digital mammograms. A feature selection algorithm based the on null-hypothesis test with statistical 'two-sample t-test' method has been suggested to select most significant coefficients from a large number of DOST coefficients. The selected coefficients are used as features in the classification of mammographic images as benign or malignant. This scheme utilises an AdaBoost algorithm with random forest as its base classifier. Two standard databases Mammographic Image Analysis Society (MIAS) and Digital Database for Screening Mammography (DDSM) are used for the validation of the proposed scheme. Simulation results show an optimal classification performance with respect to accuracies of 98.3 and 98.8% and AUC (receiver operating characteristic) values of 0.9985 and 0.9992 for MIAS and DDSM, respectively. Comparative analysis shows that the proposed scheme outperforms its competent schemes.

Developing and applying novel spectral feature parameters for classifying soil salt types in arid land

Soil salinization is a major desertification process that threatens especially the stability of arid ecosystems. There is an urgent need for intensive monitoring and quick assessment of salinization through remote sensing as a tool for combating desertification in such ecosystems. Recent researches have revealed that in order to retrieve soil salt contents accurately from hyperspectral reflectance, a pre-knowledge of salt types is required, which greatly outlines the spectral features of saline soil reflectance. In this study, a set of feature parameters have been developed after a thorough investigation of spectral responses to different soil salt types and salt contents for quick and accurate classification of soil salt types. The application has been validated using three independent datasets composed from: laboratory experiments (dataset I), in-situ field measurements (dataset II), and satellite-borne Hyperion image (dataset III). For comparison, four other common classification algorithms have been validated using the same datasets. The results showed that the new approach proposed in this study performed well with not only single-type but also multiple-type salts for which the four common algorithms performed rather fairly. Furthermore, validating using datasets II and III showed that the newly proposed approach had a stable performance while the other four failed, indicating the advantage of the new approach. The feature parameters developed in this study hence provide a novel and efficient approach for salt type classification from reflectance spectra, and we foresee its potential applications on large-scale soil salt type mapping towards better understanding soil salinity characterization from remote sensing data.

Simple and Computationally Efficient Movement Classification Approach for EMG-controlled Prosthetic Hand: ANFIS vs. Artificial Neural Network

The aim of this paper is to propose an exploratory study on simple, accurate and computationally efficient movement classification technique for prosthetic hand application. The surface myoelectric signals were acquired from 2 muscles—Flexor Carpi Ulnaris and Extensor Carpi Radialis of 4 normal-limb subjects. These signals were segmented and the features extracted using a new combined time-domain method of feature extraction. The fuzzy C-mean clustering method and scatter plots were used to evaluate the performance of the proposed multi-feature versus other accurate multi-features. Finally, the movements were classified using the adaptive neuro-fuzzy inference system (ANFIS) and the artificial neural network. Comparison results indicate that ANFIS not only displays higher classification accuracy (88.90%) than the artificial neural network, but it also improves computation time significantly.

Classification of Scenes into Indoor/Outdoor

Effective model for scene classification is essential, to access the desired images from large scale databases. This study presents an efficient scene classification approach by integrating low level features, to reduce the semantic gap between the visual features and richness of human perception. The objective of the study is to categorize an image into indoor or outdoor scene using relevant low level features such as color and texture. The color feature from HSV color model, texture feature through GLCM and entropy computed from UV color space forms the feature vector. To support automatic scene classification, Support Vector Machine (SVM) is implemented on low level features for categorizing a scene into indoor/outdoor. Since the combination of these image features exhibit a distinctive disparity between images containing indoor or outdoor scenes, the proposed method achieves better performance in terms of classification accuracy of about 92.44%. The proposed method has been evaluated on IITM- SCID2 (Scene Classification Image Database) and dataset of 3442 images collected from the web.

Principal Association Mining: An efficient classification approach

Classification is one of the key tasks in business intelligence, decision science, and machine learning. Associative classification has aroused significant research interest in recent years due to its superior accuracy. Traditional association rule mining algorithms often yield many redundant and sometimes conflicting class association rules. This paper presents a new, efficient associative classification approach. This new approach produces a compact classifier with a small number of association rules, yet with good classification performance. This approach is based on a novel rule quality metric, named as Principality, which measures an association rule’s classification accuracy and coverage for a specific class. Heuristic methods utilizing the Principality metric are applied to rule pruning and associative classifier construction to produce a compact classifier. This Principal Association Mining (PAM) approach is confirmed to be effective at improving classification accuracy as well as decreasing classifier size by experiments conducted on 17 datasets.

Fly Wing Biometrics Using Modified Local Binary Pattern, SVMs and Random Forest

This paper presents an efficient approach for classification of the gender of a common fruit fly, Drosophila melanogaster, based on their wing's texture. The novelty of this research effort is that a Modified Local Binary Pattern (MLBP), which combines both the sign and magnitude features for the improvement of fly wing's texture classification performance, is applied. The extracted features are then used to classify the gender of the fruit fly by using the Support Vector Machines (SVMs) and Random Forest (RF). We validate the performance of the proposed scheme on two fly wing datasets. The highest accuracy achieved by the proposed approach is 94%. In this paper, we limit our approach to gender classification; however, this effort can be extended to explore important characteristics of a fly using wing's texture analysis.

An Efficient Approach for Medical Image Classification using Association Rules

images are crucial in diagnosis, therapy, surgery, reference, and training. The availability of Digital radiology equipment availability has ensured that digital medical images management gets more attention now. This paper presents an automatic classification system for Computed Tomography (CT) medical images are presented in this paper. In the presented methodology, Bi-orthogonal spline wavelet was used to extract features from the brain, chest and colon CT scan images. Association Rule Mining (ARM) was used for feature reduction leading to the selection of attributes with respect to class label based on frequent sets. On feature selection the CT images are classified using Naive Bayes and k-Nearest Neighbor algorithm. The classification accuracy obtained was compared with and without feature selection using Association Rule Mining.

An efficient approach for feature extraction and classification of microarray cancer data

DNA microarray consists of huge amount of features with small number of samples. In this paper, we address the dimension reduction of DNA features in which relevant features are extracted among thousands of irrelevant ones through dimensionality reduction. This enhances the speed and accuracy of the classifiers. Principal component analysis (PCA) is a very powerful statistical technique, is used to satisfy the aim, is to project the original I-dimensional space into an I

An Adaptive Machine Learning Approach for Semantic Analysis to Extract Medical Knowledge

Machine learning approach over medical datasets is still an important research issue in recent days of technology in medical field. In our approach we are proposing an efficient classification approach for analysis of testing samples with training samples. Initially we train the medical abstracts by identifying disease and treatment and then forwards these informative and non informative sentences towards word of bag (Cure, prevent, etc.) to extract positive and negative sentences and finally update them in database for future classification of testing samples.