Classification Success Rate Research Articles

Cut-edge detection method for wheat harvesting based on stereo vision

A cut-edge detection method for wheat based on stereo vision was proposed in this work to obtain the navigation path of a combine harvester. First, the point cloud was acquired by the stereo camera. The crop area was extracted with the threshold obtained by the Otsu method. Then, the point cloud of the crop area was gridded. The grids were clustered by the density-based spatial clustering of applications with noise method to classify the different crop areas. After filtering the noise caused by the ridge in the yield, the grids of interest were extracted. The edge point was extracted in each grid of interest. The polynomial fitting method was then used to acquire the straight or curved cut-edge. A total of 300 images were selected for the test of crop area extraction and crop areas classification. The results showed that the success rate of crop area extraction was 93.7% and the success rate of crop areas classification was 91.1%. 100 images were selected to extract the edge points and compare with the true value of manual measurement. Experiment results showed that the average deviation of the edge points was 8.47 cm, the maximum deviation was 23.1 cm, and the standard deviation was 5.97 cm. The proposed method is thus capable of providing support for the automatic navigation of combine harvesters.

Comparison of Multilayer Neural Network Models in Terms of Success of Classifications Based on EmguCV, ML.NET and Tensorflow.Net

In this paper, we compare three different models of multilayer neural networks in terms of their success in the classification phase. These models were designed for EmguCV, ML.NET and Tensorflow.Net libraries, which are currently among the most widely used libraries in the implementation of an automatic recognition system. Using the EmguCV library, we achieved a success rate in the classification of human faces of 81.95% and with ML.NET, which was based on the pre-trained ResNet50 model using convolution layers, up to 91.15% accuracy. The result of the success of the classification process was influenced by the time required for training and also the time required for the classification itself. The Tensorflow.Net model did not show sufficient classification ability when classifying using vector distances; the highest success rate of classification was only 13.31%. Neural networks were trained on a dataset with 1454 photographs of faces involving 43 people. At a time when neural networks are becoming more and more used for applications of different natures, it is necessary to choose the right model in the classification process that will be able to achieve the required accuracy with the minimum time required for training. The application created by us allows the insertion of images and the creation of their own datasets, on the basis of which the user can train a model with its own parameters. Models can then be saved and integrated into other applications.

Channel based epilepsy seizure type detection from electroencephalography (EEG) signals with machine learning techniques

Epileptic seizures result from disturbances in the electrical activity of the brain, classified as focal, generalized, or unknown. Failure to correctly classify epileptic seizures may result in inappropriate treatment and continuation of seizures. Therefore, automatic detection of generalized, focal, and other epileptic seizures from EEG signals is important. In this research article, Focal-Generalized classification method is proposed that compares traditional classification algorithms and deep learning methods. Two different classifications: four-class (Case (I) Complex Partial Seizure (CPSZ) (C4-T4 Onset)-CPSZ (FP2-F8 Onset)-CPSZ (T5-O1 Onset)- Absence Seizure (ABSZ)) and two-class (Case (II) CPSZ-ABSZ) problems are considered. This study includes preprocessing of scalp Electroencephalogram (EEG) data, feature extraction with discrete wavelet method, feature selection using Correlation-based Feature Selection (CFS) method, and classification of data with classifier algorithms (K-Nearest Neighbors (Knn), Support Vector Machine (SVM), Random Forest (RF) and Long Short-Term Memory (LSTM). The proposed method was applied on 23 subjects in the Temple University Hospital (TUH) scalp EEG data set, and a classification success rate of 95,92% for case (I) and 98,08% for case (II) was successfully achieved with deep learning architecture LSTM.

Utilization of Machine Learning for the Differentiation of Positional NPS Isomers with Direct Analysis in Real Time Mass Spectrometry.

The differentiation of positional isomers is a well established analytical challenge for forensic laboratories. As more novel psychoactive substances (NPSs) are introduced to the illicit drug market, robust yet efficient methods of isomer identification are needed. Although current literature suggests that Direct Analysis in Real Time–Time-of-Flight mass spectrometry (DART-ToF) with in-source collision induced dissociation (is-CID) can be used to differentiate positional isomers, it is currently unclear whether this capability extends to positional isomers whose only structural difference is the precise location of a single substitution on an aromatic ring. The aim of this work was to determine whether chemometric analysis of DART-ToF data could offer forensic laboratories an alternative rapid and robust method of differentiating NPS positional ring isomers. To test the feasibility of this technique, three positional isomer sets (fluoroamphetamine, fluoromethamphetamine, and methylmethcathinone) were analyzed. Using a linear rail for consistent sample introduction, the three isomers of each type were analyzed 96 times over an eight-week timespan. The classification methods investigated included a univariate approach, the Welch t test at each included ion; a multivariate approach, linear discriminant analysis; and a machine learning approach, the Random Forest classifier. For each method, multiple validation techniques were used including restricting the classifier to data that was only generated on one day. Of these classification methods, the Random Forest algorithm was ultimately the most accurate and robust, consistently achieving out-of-bag error rates below 5%. At an inconclusive rate of approximately 5%, a success rate of 100% was obtained for isomer identification when applied to a randomly selected test set. The model was further tested with data acquired as a part of a different batch. The highest classification success rate was 93.9%, and error rates under 5% were consistently achieved.

Classification of Invoice Images By Using Convolutional Neural Networks

Today, as the companies grow, the number of personnel working within the company and the number of supplier companies that the company works with are also increasing. In parallel with this increase, the amount of expenditure made on behalf of the company increases, and more invoices are created. Since the invoices must be kept for legal reasons, physical invoices are transferred to the digital environment. Since large companies have large numbers of invoices, labor demand is higher in digitalizing invoices. In addition, as the number of invoices to be transferred to digital media increases, the number of possible errors during entry becomes more. This paper aims to automate the transfer of invoices to the digital environment. In this study, invoices belonging to four different templates were used. Invoice images taken from a bank system were used for the first time in this study, and the original invoice dataset was prepared. Furthermore, two more datasets were obtained by applying preprocessing methods (Zero-Padding, Brightness Augmentation) on the original dataset. The Invoice classification system developed using Convolutional Neural Networks (CNN) architectures named LeNet-5, VGG-19, and MobileNetV2 was trained on three different data sets. Data preprocessing techniques such as correcting the curvature and aspect ratio of the invoices and image augmentation with variable brightness ratio were applied to create the data sets. The datasets created with preprocessing techniques have increased the classification success of the proposed models. With this proposed model, invoice images were automatically classified according to their templates using CNN architectures. In experimental studies, a classification success rate of 99.83% was achieved in training performed on the data set produced by the data augmentation method.

Effectiveness of Human-Artificial Intelligence Collaboration in Cephalometric Landmark Detection.

Detection of cephalometric landmarks has contributed to the analysis of malocclusion during orthodontic diagnosis. Many recent studies involving deep learning have focused on head-to-head comparisons of accuracy in landmark identification between artificial intelligence (AI) and humans. However, a human–AI collaboration for the identification of cephalometric landmarks has not been evaluated. We selected 1193 cephalograms and used them to train the deep anatomical context feature learning (DACFL) model. The number of target landmarks was 41. To evaluate the effect of human–AI collaboration on landmark detection, 10 images were extracted randomly from 100 test images. The experiment included 20 dental students as beginners in landmark localization. The outcomes were determined by measuring the mean radial error (MRE), successful detection rate (SDR), and successful classification rate (SCR). On the dataset, the DACFL model exhibited an average MRE of 1.87 ± 2.04 mm and an average SDR of 73.17% within a 2 mm threshold. Compared with the beginner group, beginner–AI collaboration improved the SDR by 5.33% within a 2 mm threshold and also improved the SCR by 8.38%. Thus, the beginner–AI collaboration was effective in the detection of cephalometric landmarks. Further studies should be performed to demonstrate the benefits of an orthodontist–AI collaboration.

Hologram classification of occluded and deformable objects with speckle noise contamination by deep learning.

Advancements in optical, computing, and electronic technologies have enabled holograms of physical three-dimensional (3D) objects to be captured. The hologram can be displayed with a spatial light modulator to reconstruct a visible image. Although holography is an ideal solution for recording 3D images, a hologram comprises high-frequency fringe patterns that are almost impossible to recognize with traditional computer vision methods. Recently, it has been shown that holograms can be classified with deep learning based on convolution neural networks. However, the method can only achieve a high success classification rate if the image represented in the hologram is without speckle noise and occlusion. Minor occlusion of the image generally leads to a substantial drop in the success rate. This paper proposes a method known as ensemble deep-learning invariant occluded hologram classification to overcome this problem. The proposed new method attains over 95% accuracy in the classification of holograms of partially occluded handwritten numbers contaminated with speckle noise. To achieve the performance, a new augmentation scheme and a new enhanced ensemble structure are necessary. The new augmentation process includes occluded objects and simulates the worst-case scenario of speckle noise.

Study on the classification problem of the coping stances in the Satir model based on machine learning

ABSTRACT This paper applies machine learning technology to the Satir theory model and intelligently classifies the communication stances of the second layer according to the language and behaviour information of the first layer. We arranged a large number of dialogical language materials from a TV interview programme and used the ICTCLAS Chinese word segmentation system to create a ‘psychological consultation database’. We construct the word training set by part of making use of speech filtering and text word vectorisation, and construct the semantic training set by annotating the original data with the Satir model. These two sets form the Satir communication posture classification training set. Experimental results show that the success rate of classification of four inconsistent coping stances reached 70.37%, 75.92%, 83.33%, and 77.78%.

Population Structure of Indian Mackerel (Rastrelliger kanagurta) in Java and Bali Island, Indonesia Inferred from Otolith Shape

The Indian mackerel, Rastrelliger kanagurta, is a commercially important resource across Indonesia's archipelagic waters. Despite its essential role in the fisheries industry, information regarding its population structures for management purposes is still limited. The present study investigates the variability of otolith shape as an efficient tool for stock discrimination. A total of 159 pairs of sagittal otolith were sampled at four localities along the southern Java-Bali coastal waters. Otolith outline was modeled by using Wavelet coefficients and was compared between localities using ANOVA-like permutation test, Canonical Analysis of Principal Coordinates (CAP), and classification test performed using linear discriminant analysis. The results showed significant differences in the shape of otolith between populations, reflected explicitly in the morphological feature of the excisura major. The differences in otolith shape were examined from all localities (ANOVA-like, p<0.001, and CAP p>0.05), except those otoliths between Palabuhanratu and Pacitan, thus, a mixed stock may occur in these two locations. These findings were also confirmed by the low classification success rate using LDA 44.26%. These results suggested that at least two distinct stocks potentially contributed to the fishery, impacting species management and conservation.

Monitoring based on Analytical Redundancy and Classification for a Primary Flight Surface Electromechanical Actuator

Our objective is to develop a monitoring algorithm that achieves early detection of degradations that may lead to jamming or excessive backlash in the electromechanical actuators (EMAs) used for primary fight control. To do so, a detailed EMA model is determined in order to construct a simulator able to reproduce test bench results and simulate faulty behaviours. Exploiting the analytical redundancy within the dynamical model of the EMA, a residual generator is designed to be sensitive to a potential friction change or backlash increase, by using the typical measurement sequences associated to pre-flight test profiles. A classifier based on a multiclass Support Vector Machine approach is used to detect and identify the nature of the faults. The classifier validity is corroborated by a K-fold cross-validation. Finally, tests are conducted on the simulator under various friction and backlash situations. The tests are performed under faults of realistic amplitude while considering uncertainties in the model parameters. The developed algorithm ensures the identification of the change in the friction and/or backlash with a successful classification rate of 98.13%.

Estimation of Parameters of Finite Mixture of Rayleigh Distribution by the Expectation‐Maximization Algorithm

In the lifetime process in some systems, most data cannot belong to one single population. In fact, it can represent several subpopulations. In such a case, the known distribution cannot be used to model data. Instead, a mixture of distribution is used to modulate the data and classify them into several subgroups. The mixture of Rayleigh distribution is best to be used with the lifetime process. This paper aims to infer model parameters by the expectation‐maximization (EM) algorithm through the maximum likelihood function. The technique is applied to simulated data by following several scenarios. The accuracy of estimation has been examined by the average mean square error (AMSE) and the average classification success rate (ACSR). The results showed that the method performed well in all simulation scenarios with respect to different sample sizes.

Classification of epileptic seizures from electroencephalogram (EEG) data using bidirectional short-term memory (Bi-LSTM) network architecture

Epilepsy is a complex disease, difficult to detect and common among neurological diseases. The separation of epileptic and non-epileptic activity and the identification of the form of the epileptic activity on the electroencephalogram (EEG) plays an essential role in providing correct treatment to patients. At the same time, detection of epileptic seizures is very important for patients. In this study, a method to classify EEG data using deep neural network architecture is presented. In the presented method, bidirectional long-term memory (Bi-LSTM), a type of recurrent neural network (RNN) was used. Single-layer Bi-LSTM architecture memory units, neural numbers and learning algorithms were modeled. After determining the architectural hyperparameters, the single layer 10-neurons Bi-LSTM approach is proposed as the optimum model. In this model, the effect of both optimization algorithms on classification success and changing the number of neurons on performance was investigated. In the classification between epileptic activity and other activities, the success rate was 96%, success rate in binary classification 99% and the mean success rate for classification was 97.78%. Unlike previous work, this study did not just focus on success rates. In addition, the importance of determining the hyperparameters (selection of optimization algorithms, initial learning rate, number of neurons) in the Bi-LSTM model was highlighted and the effects of these choices on accuracy rates was determined. Success rate and processing load were examined in detail. This study, which emphasizes the importance of parameter optimization, provides an important perspective to this research area.

Classification of hazelnut varieties by using artificial neural network and discriminant analysis

Aim of study: This study was conducted to classify hazelnut (Corylus avellana L.) varieties by using artificial neural network and discriminant analysis. Area of study: Samsun Province, Turkey. Material and methods: The physical, mechanical and optical properties of 11 hazelnut varieties were determined for three major axes. The parameters of physical, mechanical and optical properties were included as independent variables, while hazelnut varieties were included as dependent variables. Models were created for each of the three axes to classify hazelnut varieties. Main results: Classification success rates with Artificial Neural Networks (ANN) and Discriminant Analysis (DA) were found as 89.1% and 92.7% for X axis, as 92.7% and 92.7% for Y axis and as 86.8% and 88.7% for Z axis, respectively. The classification results of ANN and DA models were found to be very close to each other. Both models can be used in the classification of hazelnut varieties. Research highlights: The results obtained for the identification and classification of hazelnut varieties show the feasibility and effectiveness of the proposed models.

Multi-Layer Hybrid Fuzzy Classification Based on SVM and Improved PSO for Speech Emotion Recognition

Speech Emotion Recognition (SER) plays a significant role in the field of Human–Computer Interaction (HCI) with a wide range of applications. However, there are still some issues in practical application. One of the issues is the difference between emotional expression amongst various individuals, and another is that some indistinguishable emotions may reduce the stability of the SER system. In this paper, we propose a multi-layer hybrid fuzzy support vector machine (MLHF-SVM) model, which includes three layers: feature extraction layer, pre-classification layer, and classification layer. The MLHF-SVM model solves the above-mentioned issues by fuzzy c-means (FCM) based on identification information of human and multi-layer SVM classifiers, respectively. In addition, to overcome the weakness that FCM tends to fall into local minima, an improved natural exponential inertia weight particle swarm optimization (IEPSO) algorithm is proposed and integrated with fuzzy c-means for optimization. Moreover, in the feature extraction layer, non-personalized features and personalized features are combined to improve accuracy. In order to verify the effectiveness of the proposed model, all emotions in three popular datasets are used for simulation. The results show that this model can effectively improve the success rate of classification and the maximum value of a single emotion recognition rate is 97.67% on the EmoDB dataset.

Abstract 11449: Classifying Multiple Cardiac Signals Simultaneously Using Radial Basis Function Neural Networks

Introduction: Classifying multiple cardiac signals simultaneously is still a challenge for computational methods. We propose a computational algorithm that performs a multi-class classification. We use a combination of global features of the electrocardiogram (ECG) signal, where we statistically analyze the complete signal, as well as include the local features of the cardiac cycles. Methodology: 286 ECG signals obtained from the MIT-BIH Database were evaluated:18 healthy, 48 with arrhythmia, 23 with Atrial Fibrillation, 70 with apnea, 78 with supraventricular arrhythmias, 15 with congestive heart failure, 11 with epilepsy, 5 with fetal electrocardiogram, 18 with ventricular tachyarrhythmia. For each of the 286 signals, 1200 cardiac cycles from a total of one second measurement were used, characterized by 400 milliseconds before and 600 milliseconds after the R wave. We calculated variance, skewness and kurtosis from the 286 signals, extracting 1200 features from the cardiac cycles for each of the ECG signals. These hyperparameters were used as input for the Radial Basis Function Neural Networks (RBF), where 80% were for training and 20% for testing, with cross-validation of K-fold = 10. The RBF had one hidden layer with error correction by mean squared error, number of iterations equal to 1000, number of neurons equal to 25 and learning rate 0.01. Results: Each group of cardiac signals evaluated in this study had its own features and, therefore, had unique parameters for each of the different pathologies in a three dimensional plot, as illustrated in the figure. The average accuracy of the proposed method for multi-class classification of cardiac signals was 99.98 %. Conclusion: The high success rate in the multi-class classification for eleven classes of cardiac signals makes the proposed methodology an promising alternative to aid in the diagnosis of cardiac pathologies.

CNN / Bi‐LSTM‐based deep learning algorithm for classification of power quality disturbances by using spectrogram images

This paper, using an inverse signal approach, presents a novel deep learning algorithm based on a convolutional neural network (CNN) and bidirectional long short-term memory (Bi-LSTM) together with spectrograms for the classification of power quality disturbances (PQDs). The proposed method focuses on the region where the PQD event occurs, using a deep learning-based spectrogram with the aim of increasing classification success rates. In the proposed approach, the time shift of the signal relative to the pure sine signal was found and relative to this time shift, an inverse sine wave was generated according to the original signal. By collecting this sine wave together with the original signal, spectrograms of both signals were obtained and converted into red/green/blue (RGB) images that were then combined. Finally, classification was carried out via CNN/Bi-LSTM. In this context, 29 different disturbance events in both single and combined structures were used. The proposed model was applied to the disturbance events and 99.33% classification accuracy was obtained.

Çeltik yanıklığı hastalığının görüntü işleme teknikleri kullanılarak tespit edilmesi

In rice plant, accurate and timely detection of diseases helps to start agricultural practices on time and thus reduces economic losses significantly. For this purpose, image processing techniques were used to identify and classify the rice leaf blight disease ( Pyricularia oryzae Cav.). In image processing, a clustering method was used for the segmentation of the diseased part, the non-diseased part and the background. Images of rice leaf blight disease were taken both from the ground and with the aid of a drone. Levenberg-Marquardt training algorithm was preferred in artificial neural networks model. While the RMS, R 2 and error values of the test data of MEITG proposed for identification were 0.000017, 0.9999 and 0.019%, respectively, they were found as 0.000007, 0.9999 and 0.002% for MERITD. The MCITG and MCRITD models presented for classification were found to have classification success rates of 92.2 percent and 100 percent, respectively. The results obtained for the identification and classification of rice leaf blight disease show the feasibility and effectiveness of the proposed model.

Alzheimer Identification through DNA Methylation and Artificial Intelligence Techniques

A nonlinear approach to identifying combinations of CpGs DNA methylation data, as biomarkers for Alzheimer (AD) disease, is presented in this paper. It will be shown that the presented algorithm can substantially reduce the amount of CpGs used while generating forecasts that are more accurate than using all the CpGs available. It is assumed that the process, in principle, can be non-linear; hence, a non-linear approach might be more appropriate. The proposed algorithm selects which CpGs to use as input data in a classification problem that tries to distinguish between patients suffering from AD and healthy control individuals. This type of classification problem is suitable for techniques, such as support vector machines. The algorithm was used both at a single dataset level, as well as using multiple datasets. Developing robust algorithms for multi-datasets is challenging, due to the impact that small differences in laboratory procedures have in the obtained data. The approach that was followed in the paper can be expanded to multiple datasets, allowing for a gradual more granular understanding of the underlying process. A 92% successful classification rate was obtained, using the proposed method, which is a higher value than the result obtained using all the CpGs available. This is likely due to the reduction in the dimensionality of the data obtained by the algorithm that, in turn, helps to reduce the risk of reaching a local minima.

Altered Glycosylation of Human Alpha-1-Acid Glycoprotein as a Biomarker for Malignant Melanoma.

A high-resolution HILIC-MS/MS method was developed to analyze anthranilic acid derivatives of N-glycans released from human serum alpha-1-acid glycoprotein (AGP). The method was applied to samples obtained from 18 patients suffering from high-risk malignant melanoma as well as 19 healthy individuals. It enabled the identification of 102 glycan isomers separating isomers that differ only in sialic acid linkage (α-2,3, α-2,6) or in fucose positions (core, antenna). Comparative assessment of the samples revealed that upregulation of certain fucosylated glycans and downregulation of their nonfucosylated counterparts occurred in cancer patients. An increased ratio of isomers with more α-2,6-linked sialic acids was also observed. Linear discriminant analysis (LDA) combining 10 variables with the highest discriminatory power was employed to categorize the samples based on their glycosylation pattern. The performance of the method was tested by cross-validation, resulting in an overall classification success rate of 96.7%. The approach presented here is significantly superior to serological marker S100B protein in terms of sensitivity and negative predictive power in the population studied. Therefore, it may effectively support the diagnosis of malignant melanoma as a biomarker.

A DNA-derived phage nose using machine learning and artificial neural processing for diagnosing lung cancer

There is a growing interest in electronic nose-based diagnostic systems that are fast and portable. However, existing technologies are suitable only for operation in the laboratory, making them difficult to apply in a rapid, non-face-to-face, and field-suitable manner. Here, we demonstrate a DNA-derived phage nose (D2pNose) as a portable respiratory disease diagnosis system requiring no pretreatment. D2pNose was produced based on phage colour films implanted with DNA sequences from mammalian olfactory receptor cells, and as a result, it possesses the comprehensive reactivity of these cells. The manipulated surface chemistry of the genetically engineered phages was verified through a correlation analysis between the calculated and the experimentally measured reactivity. Breaths from 31 healthy subjects and 31 lung cancer patients were collected and exposed to D2pNose without pretreatment. With the help of deep learning and neural pattern separation, D2pNose has achieved a diagnostic success rate of over 75% and a classification success rate of over 86% for lung cancer based on raw human breath. Based on these results, D2pNose can be expected to be directly applicable to other respiratory diseases.