Low Classification Performance Research Articles

Revealing GLCM Metric Variations across a Plant Disease Dataset: A Comprehensive Examination and Future Prospects for Enhanced Deep Learning Applications

This study highlights the intricate relationship between Gray-Level Co-occurrence Matrix (GLCM) metrics and machine learning model performance in the context of plant disease identification. It emphasizes the importance of rigorous dataset evaluation and selection protocols to ensure reliable and generalizable classification outcomes. Through a comprehensive examination of publicly available plant disease datasets, focusing on their performance as measured by GLCM metrics, this research identified dataset_2 (D2), a database of leaf images, as the top performer across all GLCM analyses. These datasets were then utilized to train the DarkNet19 deep learning model, with D2 exhibiting superior performance in both GLCM analysis and DarkNet19 training (achieving about 91% testing accuracy) according to performance metrics such as accuracy, precision, recall, and F1-score. The datasets other than dataset_1 and 2 exhibited significantly low classification performance, particularly in supporting GLCM analysis. The findings underscore the need for transparency and rigor in dataset selection, particularly given the abundance of similar datasets in the literature and the growing trend of utilizing deep learning methods in future scientific research.

Classification in Early Fire Detection Using Multi-Sensor Nodes-A Transfer Learning Approach.

Effective early fire detection is crucial for preventing damage to people and buildings, especially in fire-prone historic structures. However, due to the infrequent occurrence of fire events throughout a building's lifespan, real-world data for training models are often sparse. In this study, we applied feature representation transfer and instance transfer in the context of early fire detection using multi-sensor nodes. The goal was to investigate whether training data from a small-scale setup (source domain) can be used to identify various incipient fire scenarios in their early stages within a full-scale test room (target domain). In a first step, we employed Linear Discriminant Analysis (LDA) to create a new feature space solely based on the source domain data and predicted four different fire types (smoldering wood, smoldering cotton, smoldering cable and candle fire) in the target domain with a classification rate up to 69% and a Cohen's Kappa of 0.58. Notably, lower classification performance was observed for sensor node positions close to the wall in the full-scale test room. In a second experiment, we applied the TrAdaBoost algorithm as a common instance transfer technique to adapt the model to the target domain, assuming that sparse information from the target domain is available. Boosting the data from 1% to 30% was utilized for individual sensor node positions in the target domain to adapt the model to the target domain. We found that additional boosting improved the classification performance (average classification rate of 73% and an average Cohen's Kappa of 0.63). However, it was noted that excessively boosting the data could lead to overfitting to a specific sensor node position in the target domain, resulting in a reduction in the overall classification performance.

Explainable deep-learning framework: decoding brain states and prediction of individual performance in false-belief task at early childhood stage.

Decoding of cognitive states aims to identify individuals' brain states and brain fingerprints to predict behavior. Deep learning provides an important platform for analyzing brain signals at different developmental stages to understand brain dynamics. Due to their internal architecture and feature extraction techniques, existing machine-learning and deep-learning approaches are suffering from low classification performance and explainability issues that must be improved. In the current study, we hypothesized that even at the early childhood stage (as early as 3-years), connectivity between brain regions could decode brain states and predict behavioral performance in false-belief tasks. To this end, we proposed an explainable deep learning framework to decode brain states (Theory of Mind and Pain states) and predict individual performance on ToM-related false-belief tasks in a developmental dataset. We proposed an explainable spatiotemporal connectivity-based Graph Convolutional Neural Network (Ex-stGCNN) model for decoding brain states. Here, we consider a developmental dataset, N = 155 (122 children; 3-12 yrs and 33 adults; 18-39 yrs), in which participants watched a short, soundless animated movie, shown to activate Theory-of-Mind (ToM) and pain networs. After scanning, the participants underwent a ToM-related false-belief task, leading to categorization into the pass, fail, and inconsistent groups based on performance. We trained our proposed model using Functional Connectivity (FC) and Inter-Subject Functional Correlations (ISFC) matrices separately. We observed that the stimulus-driven feature set (ISFC) could capture ToM and Pain brain states more accurately with an average accuracy of 94%, whereas it achieved 85% accuracy using FC matrices. We also validated our results using five-fold cross-validation and achieved an average accuracy of 92%. Besides this study, we applied the SHapley Additive exPlanations (SHAP) approach to identify brain fingerprints that contributed the most to predictions. We hypothesized that ToM network brain connectivity could predict individual performance on false-belief tasks. We proposed an Explainable Convolutional Variational Auto-Encoder (Ex-Convolutional VAE) model to predict individual performance on false-belief tasks and trained the model using FC and ISFC matrices separately. ISFC matrices again outperformed the FC matrices in prediction of individual performance. We achieved 93.5% accuracy with an F1-score of 0.94 using ISFC matrices and achieved 90% accuracy with an F1-score of 0.91 using FC matrices.

An Ultra-Low Power Wearable BMI System with Continual Learning Capabilities.

Driven by the progress in efficient embedded processing, there is an accelerating trend toward running machine learning models directly on wearable Brain-Machine Interfaces (BMIs) to improve portability and privacy and maximize battery life. However, achieving low latency and high classification performance remains challenging due to the inherent variability of electroencephalographic (EEG) signals across sessions and the limited onboard resources. This work proposes a comprehensive BMI workflow based on a CNN-based Continual Learning (CL) framework, allowing the system to adapt to inter-session changes. The workflow is deployed on a wearable, parallel ultra-low power BMI platform (BioGAP). Our results based on two in-house datasets, Dataset A and Dataset B, show that the CL workflow improves average accuracy by up to 30.36% and 10.17%, respectively. Furthermore, when implementing the continual learning on a Parallel Ultra-Low Power (PULP) microcontroller (GAP9), it achieves an energy consumption as low as 0.45mJ per inference and an adaptation time of only 21.5 ms, yielding around 25 h of battery life with a small 100 mAh, 3.7 V battery on BioGAP. Our setup, coupled with the compact CNN model and on-device CL capabilities, meets users' needs for improved privacy, reduced latency, and enhanced inter-session performance, offering good promise for smart embedded real-world BMIs.

A Model of Continual and Deep Learning for Aspect Based in Sentiment Analysis

Sentiment Analysis is a useful tool in several social and business contexts. Aspect Sentiment Classification is a subtask in Sentiment Analysis that gives information about features or aspects of people, entities, products, or services present in reviews. Different Deep Learning models have been proposed to solve Aspect Sentiment Classification focus on a specific domain such as restaurant,hotel, or laptop reviews. However, there are few proposals for creating a single model with high performance in multiple domains. The Continual Learning approach with neural networks has been used to solve aspect classification in multiple domains. However, avoid low aspect classification performance in Continual Learning is challenging. As a consequence, potential neural networkweight shifts in the learning process in different domains or datasets.In this paper, a novel Aspect Sentiment Classification approach is proposed. Our approach combines a Transformer Deep Learning technique with a Continual Learning algorithm in different domains. The input layer used is the pre‐trained model Bidirectional Encoder Representations from Transformers. The experiments show the efficacy of our proposal with 78 .

A Model of Continual and Deep Learning for Aspect Based in Sentiment Analysis

Sentiment Analysis is a useful tool in several social and business contexts. Aspect Sentiment Classification is a subtask in Sentiment Analysis that gives information about features or aspects of people, entities, products, or services present in reviews. Different Deep Learning models have been proposed to solve Aspect Sentiment Classification focus on a specific domain such as restaurant,hotel, or laptop reviews. However, there are few proposals for creating a single model with high performance in multiple domains. The Continual Learning approach with neural networks has been used to solve aspect classification in multiple domains. However, avoid low aspect classification performance in Continual Learning is challenging. As a consequence, potential neural networkweight shifts in the learning process in different domains or datasets.In this paper, a novel Aspect Sentiment Classification approach is proposed. Our approach combines a Transformer Deep Learning technique with a Continual Learning algorithm in different domains. The input layer used is the pre‐trained model Bidirectional Encoder Representations from Transformers. The experiments show the efficacy of our proposal with 78 .

Cost-sensitive continuous ensemble kernel learning for imbalanced data streams with concept drift

In stream learning, data continuously arrives over time, often at a very high rate. For imbalanced data streams with concept drift, it becomes essential to simultaneously address classification accuracy and time efficiency. However, existing algorithms currently either demand excessive computational time or exhibit relatively lower classification performance. To address these issues, in this paper, a cost-sensitive continuous ensemble kernel learning method (CCEKL) is proposed to deal with the imbalanced data streams with concept drift. Firstly, for two-class data streams, a novel misclassification cost is proposed to adapt to the changing imbalance rate. Secondly, based on the modified loss function, a continuous kernel learning method is applied to adapt to the changing data distribution. Furthermore, the two-class method is extended to multi-class classification to deal with multi-class dynamic imbalance problems. Finally, we use an ensemble approach to address the proposed algorithm sensitivity to the initial kernel width. To further improve efficiency, a parallel approach is applied to simultaneously train multi-classifiers with different initial kernel widths. Experimental results on 30 data streams reveal that, for binary classification, CCEKL achieves superior balanced accuracy in a shorter training time. For multi-class classification, CCEKL attains better balanced accuracy without requiring excessive training time, surpassing the most of the state-of-the-art baseline algorithms.

Remote Sensing Imagery Data Analysis Using Marine Predators Algorithm with Deep Learning for Food Crop Classification.

Recently, the usage of remote sensing (RS) data attained from unmanned aerial vehicles (UAV) or satellite imagery has become increasingly popular for crop classification processes, namely soil classification, crop mapping, or yield prediction. Food crop classification using RS images (RSI) is a significant application of RS technology in agriculture. It involves the use of satellite or aerial imagery to identify and classify different types of food crops grown in a specific area. This information can be valuable for crop monitoring, yield estimation, and land management. Meeting the criteria for analyzing these data requires increasingly sophisticated methods and artificial intelligence (AI) technologies provide the necessary support. Due to the heterogeneity and fragmentation of crop planting, typical classification approaches have a lower classification performance. However, the DL technique can detect and categorize crop types effectively and has a stronger feature extraction capability. In this aspect, this study designed a new remote sensing imagery data analysis using the marine predators algorithm with deep learning for food crop classification (RSMPA-DLFCC) technique. The RSMPA-DLFCC technique mainly investigates the RS data and determines the variety of food crops. In the RSMPA-DLFCC technique, the SimAM-EfficientNet model is utilized for the feature extraction process. The MPA is applied for the optimal hyperparameter selection process in order to optimize the accuracy of SimAM-EfficientNet architecture. MPA, inspired by the foraging behaviors of marine predators, perceptively explores hyperparameter configurations to optimize the hyperparameters, thereby improving the classification accuracy and generalization capabilities. For crop type detection and classification, an extreme learning machine (ELM) model can be used. The simulation analysis of the RSMPA-DLFCC technique is performed on two benchmark datasets. The extensive analysis of the results portrayed the higher performance of the RSMPA-DLFCC approach over existing DL techniques.

The Effect of Image Size on the Deep Learning Approach for Classifying Covid-19 Pneumonia

The scope of this study is related to the classification of covid-19 pneumonia which uses X-ray images of the lungs as input and is then classified into 3 categories, namely normal (healthy), pneumonia, and covid-19 pneumonia. The classification process uses a deep learning approach, which is becoming a trend nowadays because its performance is better than conventional machine learning approaches. The image dataset in this study is a collection of X-ray images of the lungs consisting of 1,525 normal images, 1,525 pneumonia images, and 1,525 Covid-19 pneumonia images. These images vary in size, but all images are larger than 1000x1000 pixels. The large size of the image makes the training time long and the computational resources required are large. Deep learning machines need to reduce the size of the image to make the process more effective. However, changes in image size may affect classification performance. Therefore, we conducted experiments in this study to determine the effect of image size changes on the classification performance of COVID-19 pneumonia using a deep learning approach. We use the Convolutional Neural Network method, which uses 22 layers with 5 convolution layers, for classification. The experimental results show that the size of the input image affects the classification performance of covid-19 pneumonia. This study provides important information that larger image sizes do not always result in better classification performance. In addition, the image size that is too small causes very low classification performance. The performance of the classification of covid-19 pneumonia using the CNN method shows the most optimal results at an input image size of 500x500 pixels.

HDQNN-Net: An Optimal Asthma Disease Detection Technique for Voice Signal Using Hybrid Deep Q-Neural Networks

Recently, asthma patients are severely suffering COVID-19 disease, thus the asthma has become one of the dangerous diseases in the world. Further, asthma is occurring in all age groups, which causing huge loss to patient’s health. The primary way to detect the asthma in humans is done by their speech signals, as the asthma severity is increases, which manipulates the properties of speech signal. The conventional methods are failed to extract the maximum features from the speech signals, which resulted in low classification performance. Thus, this article is focused on implementation of real time asthma disease detection and identification technique from speech signals using Hybrid Deep Q Neural Networks (HDQNN). Initially, the features from the speech signals are extracted by using Krill herd optimization (KHO) approach, which extracts the detailed disease specific features. Further, the optimal features are extracted by using chaotic opposition krill herd optimization (COKHO) algorithm. Then, HDQNN is used to classify the type of asthma such as normal, and stridor classes. Further, COKHO is also used to optimize the losses generated in the HDQNN model. The simulation results shows that the proposed HDQNN method resulted in superior performance as compared to state of art approaches.

Real Time Asthma Disease Detection and Identification Technique from Speech Signals Using Hybrid Dense Convolutional Neural Network

Recently, asthma patients are severely suffering COVID-19 disease, thus the asthma has become one of the dangerous diseases in the world. Further, asthma is occurring in all age groups, which causing huge loss to patient’s health. The primary way to detect the asthma in humans is done by their speech signals, as the asthma severity is increases, which manipulates the properties of speech signal. The conventional methods are failed to extract the maximum features from the speech signals, which resulted in low classification performance. Thus, this article is focused on implementation of real time asthma disease detection and identification technique from speech signals using Multi-Feature Extraction, Selection with Hybrid Classifiers (MFESHC). Initially, speech signals are preprocessed by using Maximum likelihood estimation based spread spectrum analysis (MLE-SSA) method. Then, Improved prefix Beam Search (IPBS) based natural language processing (NLP) method is used to extract and select the best features from the preprocessed speech signals. Then, hybrid dense convolutional neural networks (HDCNN) are used to classify the type of asthma such as normal, stridor, wheezes and rattle classes. Further, Modified Crow Search (MCS) is used to optimize the losses generated in the HDCNN model. The simulation results shows that the proposed MFESHC method resulted in superior performance as compared to state of art approaches because the MCS effectively reduced the losses in the model.

The application of hybrid feature based on local mean decomposition for motor imagery electroencephalogram signal classification

AbstractThis paper proposed a hybrid feature extraction algorithm based on local mean decomposition (LMD), which has better solved the existing problems of low classification performance and adaptability limitation. LMD is employed to decompose the electroencephalogram (EEG) signal into multiple components, and then, the hybrid features based on instantaneous energy, fuzzy entropy, and mathematical morphological features are extracted on specific components, and the optimal feature combination is selected by analysis of variance (ANOVA). Finally, the classification result is output by the linear discriminant analysis (LDA) classifier. The results show that the maximum accuracy of the subjects in Data Set III of BCI‐II by the method in this paper is 92.14%, and the maximum mutual information value is 0.8. The number of novel features used in this paper is small, and the complexity of the algorithm is reduced. It can adaptively select effective features according to individual differences and has good robustness.

An Environmental Pattern Recognition Method for Traditional Chinese Settlements Using Deep Learning

The recognition of environmental patterns for traditional Chinese settlements (TCSs) is a crucial task for rural planning. Traditionally, this task primarily relies on manual operations, which are inefficient and time consuming. In this paper, we study the use of deep learning techniques to achieve automatic recognition of environmental patterns in TCSs based on environmental features learned from remote sensing images and digital elevation models. Specifically, due to the lack of available datasets, a new TCS dataset was created featuring five representative environmental patterns. We also use several representative CNNs to benchmark the new dataset, finding that overfitting and geographical discrepancies largely contribute to low classification performance. Consequently, we employ a semantic segmentation model to extract the dominant elements of the input data, utilizing a metric-based meta-learning method to enable the few-shot recognition of TCS samples in new areas by comparing their similarities. Extensive experiments on the newly created dataset validate the effectiveness of our proposed method, indicating a significant improvement in the generalization ability and performance of the baselines. In sum, the proposed method can automatically recognize TCS samples in new areas, providing a powerful and reliable tool for environmental pattern research in TCSs.

Explainable fMRI-based brain decoding via spatial temporal-pyramid graph convolutional network.

Brain decoding, aiming to identify the brain states using neural activity, is important for cognitive neuroscience and neural engineering. However, existing machine learning methods for fMRI-based brain decoding either suffer from low classification performance or poor explainability. Here, we address this issue by proposing a biologically inspired architecture, Spatial Temporal-pyramid Graph Convolutional Network (STpGCN), to capture the spatial-temporal graph representation of functional brain activities. By designing multi-scale spatial-temporal pathways and bottom-up pathways that mimic the information process and temporal integration in the brain, STpGCN is capable of explicitly utilizing the multi-scale temporal dependency of brain activities via graph, thereby achieving high brain decoding performance. Additionally, we propose a sensitivity analysis method called BrainNetX to better explain the decoding results by automatically annotating task-related brain regions from the brain-network standpoint. We conduct extensive experiments on fMRI data under 23 cognitive tasks from Human Connectome Project (HCP) S1200. The results show that STpGCN significantly improves brain-decoding performance compared to competing baseline models; BrainNetX successfully annotates task-relevant brain regions. Post hoc analysis based on these regions further validates that the hierarchical structure in STpGCN significantly contributes to the explainability, robustness and generalization of the model. Our methods not only provide insights into information representation in the brain under multiple cognitive tasks but also indicate a bright future for fMRI-based brain decoding.

An improved multi-source domain adaptation network for inter-subject mental fatigue detection based on DANN.

Electroencephalogram (EEG) is often used to detect mental fatigue because of its real-time characteristic and objective nature. However, because of the individual variability of EEG among different individuals, tedious and time-consuming calibration sessions are needed. Therefore, we propose a multi-source domain adaptation network forinter-subject mental fatigue detection named FLDANN, which is short for focal loss based domain-adversarial training of neural network. As for mental state feature extraction, power spectrum density is extracted based on the Welch method from four sub-bands of EEG signals. The features of the source domain and target domain are fed into the FLDANN network. The contributions of FLDANN include: (1) It uses the idea of adversarial to reduce feature differences between the source and target domain. (2) A loss function named focal loss is used to assign weights to source and target domain samples. The experiment result shows that when the number of the source domains increases, the classification accuracy of domain-adversarial training of neural network (DANN) gradually decreases and finally tends to be stable. The proposed method achieves an accuracy of 84.10%±8.75% on the SEED-VIG dataset and 65.42%±7.47% on the self-designed dataset. In addition, the proposed method is compared with other domain adaptation methods and the results show that the proposed method outperforms those state-of-the-art methods. The result proves that the proposed method is able to solve the problem of individual differences across subjects and to solve the problem of low classification performance of multi-source domain transfer learning.

Tri-CNN: A Three Branch Model for Hyperspectral Image Classification

Hyperspectral Image (HSI) classification methods that use Deep Learning (DL) have proven to be effective in recent years. In particular, Convolutional Neural Networks (CNNs) have demonstrated extremely powerful performance in such tasks. However, the lack of training samples is one of the main contributors to low classification performance. Traditional CNN-based techniques under-utilize the inter-band correlations of HSI because they primarily use 2D-CNNs for feature extraction. Contrariwise, 3D-CNNs extract both spectral and spatial information using the same operation. While this overcomes the limitation of 2D-CNNs, it may lead to insufficient extraction of features. In order to overcome this issue, we propose an HSI classification approach named Tri-CNN which is based on a multi-scale 3D-CNN and three-branch feature fusion. We first extract HSI features using 3D-CNN at various scales. The three different features are then flattened and concatenated. To obtain the classification results, the fused features then traverse a number of fully connected layers and eventually a softmax layer. Experimental results are conducted on three datasets, Pavia University (PU), Salinas scene (SA) and GulfPort (GP) datasets, respectively. Classification results indicate that our proposed methodology shows remarkable performance in terms of the Overall Accuracy (OA), Average Accuracy (AA), and Kappa metrics when compared against existing methods.

An adaptive learning method for the fusion information of electronic nose and hyperspectral system to identify the egg quality

Data fusion technology based on the multi-sensor system can obtain the holistic properties of samples. However, multi-sensor data fusion will bring more redundant information, which will lead to low classification performance. In this work, a multi-data-fusion-attention network (MDFA-Net) is proposed, combined with the electronic nose (e-nose) and hyperspectral system to identify the egg quality. Firstly, the gas information and spectral information of eggs are obtained under different feeding conditions. Secondly, a feature adaptive learning (FAL) unit is designed to select effective information and enhance the ability of feature expression. Thirdly, based on the FAL unit, a decision network is formed to identify the fusion information of e-nose and hyperspectral system. Finally, compared with other deep learning network models, the accuracy of MDFA-Net is 99.88%, the precision is 99.87%, the recall is 99.88%, and the F 1 -score is 99.90%, which shows better classification performance and stability. • E-nose and hyperspectral system are applied to gain the quality information of egg. • An FAL unit is proposed to fuse the gas and spectral information. • Attention mechanism is proposed to adaptively focus on important features. • An MDFA-Net is proposed to achieve the decision of fusion information.

Feature Extraction Method Based on Sparse Autoencoder for Air Traffic Management System Security Situation Awareness

In wide-area distributed scenarios, it is particularly important to carry out information security situational awareness for the air traffic management (ATM) system with integrated air-ground structure. The operation data of the communication, navigation and surveillance (CNS) equipment of ATM system have the characteristics of multi-dimension, complexity, and strong correlation. In the process of situation awareness feature extraction, there are problems such as poor model accuracy, weak feature expression ability, and low classification performance. A feature association algorithm is designed to solve the above problems. Based on this algorithm, a deep-related sparse autoencoder (DRSAE) model based on improved sparse autoencoder is established. In DRSAE model, L1 regularization and Kullback–Leibler divergence (KLD) sparsity terms are used to penalize the parameters of the encoder network, and the quantity of hidden layers is increased to allow the model to optimize the global encoder network by iteratively training a single encoder. Moreover, the proposed DRSAE model and other feature extraction models such as principal component analysis (PCA), autoencoder (AE), and sparse autoencoder (SAE) are compared and evaluated by using the support vector machine (SVM) classifier. Compared with other feature extraction models, it is found that the proposed DRSAE model has good robustness in feature extraction of ATM system, and the obtained features have strong expression ability, which enhances the classification performance of the model and is convenient for situation awareness.

Online social network individual depression detection using a multitask heterogenous modality fusion approach

In recent years, the number of people who endanger their lives has been increasing rapidly due to the mental burden of depression. The online social network (OSN) provides researchers with another perspective for detecting individuals suffering from depression. However, existing machine learning-based depression detection studies still leave relatively low classification performance, suggesting that there is significant improvement potential in their feature engineering. In this paper, we manually build and publish a large dataset on Sina Weibo (a leading OSN with the largest number of active users in the Chinese community), namely the Weibo User Depression Detection Dataset (WU3D). It includes more than 20,000 normal users and more than 10,000 depressed users, both of which are labeled and rechecked following the DSM-5 official medical and psychological depression document by professionals. Then, we conclude and propose ten statistical features by analyzing the user’s text, social behavior, and posted pictures. In the meantime, text-based word features are extracted using the popular pretrained model XLNet. Moreover, we fuse these features from heterogeneous modalities and implement a multitask learning scheme to train our proposed deep neural network classification model, i.e. FusionNet (FN). The experimental results show that FN has excellent to recognize depressed users on the OSN, achieving the highest F1-Score of 0.9772 on the test set. Compared to existing studies, the proposed method has better classification performance and robustness for unbalanced training samples, as well as reasonable training and inference time. Our work provides a method to fuse multimodal information to detect individual-level depression and has reference significance for similar studies on other OSNs.

EXtreme Gradient Boosting-based method to classify patients with COVID-19

Different demographic, clinical and laboratory variables have been related to the severity and mortality following SARS-CoV-2 infection. Most studies applied traditional statistical methods and in some cases combined with a...