Improve Classification Accuracy Research Articles

Statistical and Mathematical Approaches to Disease Identification in Pennisetum glaucum Using Hybrid Optimization Techniques

To achieve maximum production and food security in the realm of agricultural biotechnology, timely and efficient disease detection in crops such as pearl millet, or Pennisetum glaucum, is crucial. This study explores the application of hybrid optimization methods to enhance the identification of Pennisetum glaucum disease. Traditional sickness classification techniques frequently have low accuracy and efficacy, necessitating research into more advanced techniques. Our approach integrates a variety of optimization techniques, including genetic algorithms (GA), particle swarm optimization (PSO), and artificial neural networks (ANN), to create a dependable hybrid model. The hybrid model enhances classification performance overall by utilizing the benefits of each unique strategy. We collected a sizable dataset of samples of Pennisetum glaucum suffering from various illnesses in order to train and validate our model. The results demonstrate a significant improvement in sickness classification accuracy when compared to conventional methods, with the hybrid model achieving a precision rate of more than 95%. Furthermore, hybrid optimization strategies reduced the computational time required for model training and prediction. The current study demonstrates how hybrid optimization has the potential to transform agricultural disease management techniques by offering a scalable and practical answer to farmers and agronomists.

Sentinel-1 (S1) time series alignment method for rapeseed fields mapping

IntroductionThis paper presents a comprehensive analysis of rapeseed fields mapping using Sentinel-1 (S1) time series data. We applied a time series alignment method to enhance the accuracy of rapeseed fields detection, even in scenarios where reference label data are limited or not available.MethodsTo this end, for five different study sites in France and North America, we first investigated the temporal transferability of the classifiers across several years within the same site, specifically using the Random Forest (RF) and InceptionTime algorithms. We then examined the spatiotemporal transferability of the classifiers when a classifier trained on one site and year was used to generate rapeseed fields map for another site and year. Next, we proposed an S1 time series alignment method to improve classification accuracy across sites and years by accounting for temporal shifts caused by differences in agricultural practices and climatic conditions between sites.Results and discussionThe main results demonstrated that rapeseed detection for 1 year, using training data from another year within the same site, achieved high accuracy, with F1 scores ranging from 85.5% to 97% for RF and from 88.2% to 98.3% for InceptionTime. When classifying using one-year training data from one site to classify another year in a different site, F1 scores varied between 48.8% and 97.7% for both RF and InceptionTime. Using a three year training dataset from one site to classify rapeseed fields in another site resulted in F1 scores ranging from 82.7% to 97.8% with RF and from 88.7% to 97.1% with InceptionTime. The proposed alignment method, designed to enhance classification using training and test data from different sites, improved F1 scores by up to 46.7%. These findings confirm the feasibility of mapping rapeseed with S1 images across various sites and years, highlighting its potential for both national and international agricultural monitoring initiatives.

A Comparative Analysis Between CNNs and ViTs for MRI-based Brain Tumor Classification

Brain tumor classification from Magnetic Resonance Imaging (MRI) images is an important task in medical imaging for the determination of appropriate treatment strategies and improvement in patient outcomes. Brain tumors, including gliomas, meningiomas, and glioblastomas, are of the most lethal forms of cancer. This research explored the potential of replacing convolutional neural networks (CNNs) by Vision Transformers (ViTs) on classifying brain tumors by MRI images. The paper focused on pretrained model Vit-B16, comparing it with traditional CNNs, including VGG16, ResNet-50, and EfficientNet-B0. ViT-B16, pretrained on ImageNet-21k, achieves improved accuracy, precision, recall, and F1-Score after being fine-tuned on the brain tumor dataset when applying data augmentation. The self-attention mechanism helps ViTs capture long-range dependencies and global context from the images, significantly improving the performance. As shown in the results, ViTs can efficiently handle complex dataset and become a useful tool for the area of medical imaging classification. This paper emphasizes the potential of Vision transformers in improving classification accuracy in the diagnosis of brain tumors. Future work can be focused on exploring better ViT architectures and data augmentation techniques.

Development of a Deep Mask Region-Based Convolutional Neural Network with Improved Weighted Quantum Wolf Optimization for Sickle Cell Anaemia Detection and Classification

To effectively manage Sickle Cell Anaemia (SCA), a severe inherited blood condition, early detection and accurate categorization are essential. Timely and precise identification is hindered by the high error rates, limited scalability, and dependency on personal expertise associated with conventional diagnostic approaches. To address these challenges, this study proposes a novel Deep Mask Region-Based Convolutional Neural Network (DMRCNN) combined with Improved Weighted Quantum Wolf Optimization (IWQWO) for SCA identification and classification. By leveraging advanced feature extraction and segmentation methods, the DMRCNN enables precise localization and identification of abnormal cells in blood smear images. The IWQWO method optimizes the DMRCNN's hyper parameters enhancing the network's efficiency by achieving an optimal balance between convergence speed and prediction accuracy. The primary objectives of this study are to improve classification accuracy, reduce computational overhead, and minimize false positives and false negatives in SCA diagnostics. Experimental results demonstrate that the proposed system achieves 96.8% precision, 97.2% accuracy, 96.5% recall and 96.8% F1-score, surpassing existing approaches with superior metrics. These findings underscore the effectiveness of the proposed method in automating SCA detection and classification, paving the way for potential integration into clinical workflows for accurate and timely assessments.

AI powered model for Oral-cancer Detection

Oral cancer detection, a critical health challenge due to late-stage diagnoses, has seen significant advancements through the integration of both hardware and software technologies. The system, designed for the early detection of oral cancer, combines innovative hardware, such as robotic arms for assistive feeding, cost-effective control systems for industrial robotics using Raspberry Pi, and Io T-enabled devices for remote oral examinations, with advanced software solutions. These hardware systems utilize lightweight structures, motorized joints, stepper motors and precise motion control to enhance functionality in healthcare and industrial environments. For early detection, Machine Learning (ML) and Deep Learning (DL) techniques, including Convolutional Neural Networks (CNNs) and transfer learning models like VGG19 and InceptionNetV3, Res-Net and Dense-Net, have been employed to analyze medical images. Image processing methods such as GLCM, wavelet transforms, and CNN-based feature extraction improve classification accuracy, while optimization techniques like Fuzzy Particle Swarm Optimization (FPSO) and Artificial Bee Colony (ABC) enhance performance. The integration of these systems achieves high sensitivity, specificity, and precision, making them effective in early oral cancer detection. By combining advanced diagnostic technologies, the system offers a comprehensive approach that improves early diagnosis, enhances patient outcomes, and complements traditional methods. Key Words: Oral Cancer, machine learning, early detection.

An Efficient Target-to-Area Classification Strategy with a PIP-Based KNN Algorithm for Epidemic Management

During a widespread epidemic, a large portion of the population faces an increased risk of contracting infectious diseases such as COVID-19, monkeypox, and pneumonia. These outbreaks often trigger cascading effects, significantly impacting society and healthcare systems. To contain the spread, the Centers for Disease Control and Prevention (CDC) must monitor infected individuals (targets) and their geographical locations (areas) as a basis for allocating medical resources. This scenario is a Target-to-Area (TTA) problem. Previous research introduced the Point-In-Polygon (PIP) technique to address multi-target and single-area TTA problems. PIP technology relies on an area’s boundary points to determine whether a target is within that region. However, when dealing with multi-target, multi-area TTA problems, PIP alone may have limitations. The K-Nearest Neighbors (KNN) algorithm presents a promising alternative, but its classification accuracy depends on the availability of sufficient samples, i.e., known targets and their corresponding geographical areas. When sample data are limited, the effectiveness of KNN is constrained, potentially delaying the CDC’s ability to track and manage outbreaks. For this problem, this study proposes an improved approach that integrates PIP and KNN technologies while introducing area boundary points as additional samples. This enhancement aims to improve classification accuracy and mitigate the impact of insufficient sample data on epidemic tracking and management.

Stress management with HRV following AI, semantic ontology, genetic algorithm and tree explainer

Heart Rate Variability (HRV) serves as a vital marker of stress levels, with lower HRV indicating higher stress. It measures the variation in the time between heartbeats and offers insights into health. Artificial intelligence (AI) research aims to use HRV data for accurate stress level classification, aiding early detection and well-being approaches. This study’s objective is to create a semantic model of HRV features in a knowledge graph and develop an accurate, reliable, explainable, and ethical AI model for predictive HRV analysis. The SWELL-KW dataset, containing labeled HRV data for stress conditions, is examined. Various techniques like feature selection and dimensionality reduction are explored to improve classification accuracy while minimizing bias. Different machine learning (ML) algorithms, including traditional and ensemble methods, are employed for analyzing both imbalanced and balanced HRV datasets. To address imbalances, various data formats and oversampling techniques such as SMOTE and ADASYN are experimented with. Additionally, a Tree-Explainer, specifically SHAP, is used to interpret and explain the models’ classifications. The combination of genetic algorithm-based feature selection and classification using a Random Forest Classifier yields effective results for both imbalanced and balanced datasets, especially in analyzing non-linear HRV features. These optimized features play a crucial role in developing a stress management system within a Semantic framework. Introducing domain ontology enhances data representation and knowledge acquisition. The consistency and reliability of the Ontology model are assessed using Hermit reasoners, with reasoning time as a performance measure. HRV serves as a significant indicator of stress, offering insights into its correlation with mental well-being. While HRV is non-invasive, its interpretation must integrate other stress assessments for a holistic understanding of an individual’s stress response. Monitoring HRV can help evaluate stress management strategies and interventions, aiding individuals in maintaining well-being.

Challenging Cognitive Load Theory: The Role of Educational Neuroscience and Artificial Intelligence in Redefining Learning Efficacy.

Background/Objectives: This systematic review integrates Cognitive Load Theory (CLT), Educational Neuroscience (EdNeuro), Artificial Intelligence (AI), and Machine Learning (ML) to examine their combined impact on optimizing learning environments. It explores how AI-driven adaptive learning systems, informed by neurophysiological insights, enhance personalized education for K-12 students and adult learners. This study emphasizes the role of Electroencephalography (EEG), Functional Near-Infrared Spectroscopy (fNIRS), and other neurophysiological tools in assessing cognitive states and guiding AI-powered interventions to refine instructional strategies dynamically. Methods: This study reviews n = 103 papers related to the integration of principles of CLT with AI and ML in educational settings. It evaluates the progress made in neuroadaptive learning technologies, especially the real-time management of cognitive load, personalized feedback systems, and the multimodal applications of AI. Besides that, this research examines key hurdles such as data privacy, ethical concerns, algorithmic bias, and scalability issues while pinpointing best practices for robust and effective implementation. Results: The results show that AI and ML significantly improve Learning Efficacy due to managing cognitive load automatically, providing personalized instruction, and adapting learning pathways dynamically based on real-time neurophysiological data. Deep Learning models such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Support Vector Machines (SVMs) improve classification accuracy, making AI-powered adaptive learning systems more efficient and scalable. Multimodal approaches enhance system robustness by mitigating signal variability and noise-related limitations by combining EEG with fMRI, Electrocardiography (ECG), and Galvanic Skin Response (GSR). Despite these advances, practical implementation challenges remain, including ethical considerations, data security risks, and accessibility disparities across learner demographics. Conclusions: AI and ML are epitomes of redefinition potentials that solid ethical frameworks, inclusive design, and scalable methodologies must inform. Future studies will be necessary for refining pre-processing techniques, expanding the variety of datasets, and advancing multimodal neuroadaptive learning for developing high-accuracy, affordable, and ethically responsible AI-driven educational systems. The future of AI-enhanced education should be inclusive, equitable, and effective across various learning populations that would surmount technological limitations and ethical dilemmas.

Acoustic-Based Industrial Diagnostics: A Scalable Noise-Robust Multiclass Framework for Anomaly Detection

This study proposes a framework for anomaly detection in industrial machines with a focus on robust multiclass classification using acoustic data. Many state-of-the-art methods only have binary classification capabilities for each machine, and suffer from poor scalability and noise robustness. In this context, we propose the use of Smoothed Pseudo Wigner–Ville Distribution-based Mel-Frequency Cepstral Coefficients (SPWVD-MFCCs) in the framework which are specifically tailored for noisy environments. SPWVD-MFCCs, with better time–frequency resolution and perceptual audio features, improve the accuracy of detecting anomalies in a more generalized way under variable signal-to-noise ratio (SNR) conditions. This framework integrates a CNN-LSTM model that efficiently and accurately analyzes spectral and temporal information separately for anomaly detection. Meanwhile, the dimensionality reduction strategy ensures good computational efficiency without losing critical information. On the MIMII dataset involving multiple machine types and noise levels, it has shown robustness and scalability. Key findings include significant improvements in classification accuracy and F1-scores, particularly in low-SNR scenarios, showcasing its adaptability to real-world industrial environments. This study represents the first application of SPWVD-MFCCs in industrial diagnostics and provides a noise-robust and scalable method for the detection of anomalies and fault classification, which is bound to improve operational safety and efficiency within complex industrial scenarios.

Security detection algorithm using CNN: Anomaly detection for API call sequence

This study proposes a security detection algorithm based on convolutional neural networks (CNNs) to enhance anomaly detection in API call sequences, addressing the challenges of capturing complex temporal relationships and nonlinear features in high-dimensional sparse API data. The proposed algorithm includes several preprocessing steps, such as deduplication to reduce redundancy, feature extraction using the TF-IDF (term frequency-inverse document frequency) algorithm, and logarithmic transformation to mitigate the impact of high-frequency APIs. An importance scoring mechanism is introduced to quantify the role of each API in anomaly detection. A customized TextCNN architecture is designed for API sequences, incorporating input layers, word embedding, multi-size convolution and pooling layers, attention layers, and fully connected layers. The attention layer is particularly applied to enhance the detection efficiency of evasion features. The model is trained using the Sigmoid activation function, CrossEntropyLoss loss function, and optimized via the Adam algorithm. The Softmax function is utilized to transform the feature vector into a probability distribution, with a threshold of 0.5 for anomaly detection. Clustering and auxiliary information are integrated to further improve classification accuracy and guide security strategy formulation. Experimental results demonstrate that the optimized TextCNN anomaly detection algorithm achieves an average accuracy of 95.88%, a recall rate of 91.23%, a false positive rate of 2.34%, and a false negative rate of 1.78%. These findings highlight the algorithm’s ability to enhance feature extraction accuracy, improve high-dimensional data processing, and provide an effective solution for real-time security monitoring, thus strengthening the security of the development environment.

Lightweight mechanical equipment fault diagnosis framework based on GCGAN-MDSCNN-ICA model

In response to the challenges posed by imbalanced failure diagnosis samples, limited labeled data, and significant computational costs in actual industrial production settings, this paper introduces a high-precision, low-resource, end-to-end fault diagnosis framework. On one hand, we propose a data augmentation method based on GCGAN, which combines CNN and GRU to construct core network structures for the generator and discriminator. We integrate a novel Smoothed Hinge-Cross-Entropy loss function to facilitate the training process, effectively mitigating mode collapse and vanishing gradient issues. On the other hand, we design a lightweight fault diagnosis model based on MDSCNN-ICA-BiGRU. By substituting standard convolutions with depthwise separable convolutions on deeper channels, the model complexity is significantly reduced, facilitating effective extraction of multiscale spatial features. The improved Coordinate Attention (CA) mechanism filters out noise and enhances the extraction of high-frequency characteristics. Combined with BiGRU, the model captures global temporal associations, achieving a fusion of spatiotemporal features. Experimental results demonstrate that the proposed approach performs well on both publicly available simulation datasets and private laboratory datasets. Compared to other benchmark methods, the GCGAN module significantly enhances data augmentation, improving classification accuracy on CNNs by 10%. When compared with classic convolutional networks such as DRSN and WDCNN, our MDSCNN-ICA-BiGRU shows faster and more stable convergence rates, with near-100% accuracy on test sets and an average computation cost reduction of approximately 70%. Even in noisy environments, our method maintains high accuracy with a slow rate of precision decay, indicating robustness and generalization capabilities.

A Novel Approach for Maize Straw Type Recognition Based on UAV Imagery Integrating Height, Shape, and Spectral Information

Accurately determining the distribution and quantity of maize straw types is of great significance for evaluating the effectiveness of conservation tillage, precisely estimating straw resources, and predicting the risk of straw burning. The widespread adoption of conservation tillage technology has greatly increased the diversity and complexity of maize straw coverage in fields after harvest. To improve the precision and effectiveness of remote sensing recognition for maize straw types, a novel method was proposed. This method utilized unmanned aerial vehicle (UAV) multispectral imagery, integrated the Stacking Enhanced Straw Index (SESI) introduced in this study, and combined height, shape, and spectral characteristics to improve recognition accuracy. Using the original five-band multispectral imagery, a new nine-band image of the study area was constructed by integrating the calculated SESI, Canopy Height Model (CHM), Product Near-Infrared Straw Index (PNISI), and Normalized Difference Vegetation Index (NDVI) through band combination. An object-oriented classification method, utilizing a “two-step segmentation with multiple algorithms” strategy, was employed to integrate height, shape, and spectral features, enabling rapid and accurate mapping of maize straw types. The results showed that height information obtained from the CHM and spectral information derived from SESI were essential for accurately classifying maize straw types. Compared to traditional methods that relied solely on spectral information for recognition of maize straw types, the proposed approach achieved a significant improvement in overall classification accuracy, increasing it by 8.95% to reach 95.46%, with a kappa coefficient of 0.94. The remote sensing recognition methods and findings for maize straw types presented in this study can offer valuable information and technical support to agricultural departments, environmental protection agencies, and related enterprises.

Enhancing Web Security through Machine Learning-based Detection of Phishing Websites

The rise of cyberattacks has led to an increase in the creation of fake websites by attackers, who use these sites for advertising products, transmit malware, or steal valuable login credentials. Phishing, the act of soliciting sensitive information from users by masquerading as a trustworthy entity, is a common technique used by attackers to achieve their goals. Spoofed websites and email spoofing are often used in phishing attacks, with spoofed emails redirecting users to phishing websites in order to trick them into revealing their personal information. Traditional solutions for detecting phishing websites rely on signature-based approaches that are not effective in detecting newly created spoofed websites. To address this challenge, researchers have been exploring machine-learning methods for detecting phishing websites. In this paper, we suggest a new approach that combines the use of blacklists and machine learning techniques such that a variety of powerful features, including domain-based features, abnormal features, and abnormal features based on URLs, HTML, and JavaScript, to rank web pages and improve classification accuracy. Our experimental results show that using the proposed approach, the random forest classifier offers the best accuracy of 93%, with FPR and FNR as 0.12 and 0.02, with a Precision of 90%, Recall of 97% an F1 Score of 93%, and MCC of 0.85.

Enhancing Automotive Safety through Context-Aware Ontology Classification

Dynamic contextual feature extraction has emerged as a critical approach for classifying unstructured automotive safety data into domain-specific ontologies. This article presents a novel framework that leverages part-of-speech tagging, positional probabilities, and optimized feature vectors to process diverse safety datasets effectively. This methodology introduces adaptive context windows and domain-aware feature extraction techniques, demonstrating marked improvements in classification accuracy compared to traditional approaches. This article shows a substantial enhancement in pattern recognition and fault classification capabilities through extensive evaluation using real-world automotive fault logs, crash reports, and service feedback data. The system particularly excelled in identifying subtle correlations between sensor anomalies and environmental conditions, while maintaining robust performance across varying data contexts. This article advances the field of automotive safety data processing by providing a scalable, context-aware solution that addresses the growing complexity of modern vehicle systems. This article suggests promising applications in real-time safety monitoring and predictive maintenance systems.

Advanced Hybrid Feature Extraction and DeepLab v3+ Semantic Segmentation for Enhanced Breast Cancer Detection with Cuckoo Search-Optimized Neural Network Classifier

This paper introduces a hybrid methodology for enhancing breast cancer detection using thermographic images. The approach combines DeepLab v3+ semantic segmentation with advanced feature extraction techniques, improving classification accuracy. DeepLab v3+ isolates regions of interest (ROIs) from thermograms, focusing on areas likely indicating cancerous lesions. Feature extraction methods like morphological statistical features, Discrete Wavelet Transform (DWT), Local Binary Patterns (LBP), and Enhanced Marine Predator Optimization (EMPO)-Optimized Dual-Tree Complex Wavelet Transform (DTCWT) capture diverse thermal patterns indicative of breast cancer. Principal Component Analysis (PCA) is used for dimensionality reduction, enhancing computational efficiency while preserving important information. The extracted features are classified using a Cuckoo Search-Optimized Neural Network (CSA-NN), which optimizes neural network parameters and addresses class imbalance and feature redundancy. Experimental results demonstrate that this hybrid methodology outperforms traditional techniques, achieving high accuracy, sensitivity, specificity, and precision across breast cancer thermography datasets. The proposed method highlights the potential of combining advanced image processing and machine learning for reliable breast cancer detection.

Bio-plausible reconfigurable spiking neuron for neuromorphic computing.

Biological neurons use diverse temporal expressions of spikes to achieve efficient communication and modulation of neural activities. Nonetheless, existing neuromorphic computing systems mainly use simplified neuron models with limited spiking behaviors due to high cost of emulating these biological spike patterns. Here, we propose a compact reconfigurable neuron design using the intrinsic dynamics of a NbO2-based spiking unit and excellent tunability in an electrochemical memory (ECRAM) to emulate the fast-slow dynamics in a bio-plausible neuron. The resistance of the ECRAM was effective in tuning the temporal dynamics of the membrane potential, contributing to flexible reconfiguration of various bio-plausible firing modes, such as phasic and burst spiking, and exhibiting adaptive spiking behaviors in changing environment. We used the bio-plausible neuron model to build spiking neural networks with bursting neurons and demonstrated improved classification accuracies over simplified models, showing great promises for use in more bio-plausible neuromorphic computing systems.

New Mitogenomes from the Genus Ablabesmyia (Diptera: Chironomidae, Tanypodiinae): Characterization and Phylogenetic Implications.

(1) Background: The insect mitogenome encodes essential genetic components and serves as an effective marker for molecular identification and phylogenetic analysis in insects due to its small size, maternal inheritance, and rapid evolution. The morphological identification of Ablabesmyia is challenging, particularly for non-experts. Thus, there is an increasing need for molecular data to improve classification accuracy and phylogenetic analysis. (2) Methods: Our analysis encompassed eight species of Ablabesmyia, a single species of Conchapelopia, one species of Denopelopia, and one species of Thienemannimyia, all originating from China. We then performed a comprehensive analysis of the nucleotide composition, sequence length, and evolutionary rate. (3) Results: All newly assembled mitogenomes displayed a negative GC-skew, indicating a cytosine bias, while most exhibited a positive AT-skew, reflecting an adenine and thymine abundance. All thirteen protein-coding genes (PCGs) featured the conventional start codon ATN, aligning closely with the typical mitochondrial start codon observed in insects. The evolutionary rates of these PCGs can be ordered as follows: ND2 > ATP8 > ND6 > ND4 > ND5 > ND3 > ND4L > ND1 > CYTB > COIII > ATP6 > COII > COI. (4) Conclusions: These newly sequenced mitogenomes exhibit structural features and nucleotide compositions that closely align with those of previously reported Chironomidae species, marking a significant expansion of the chironomid mitogenome database.

Generalizability of machine learning models for diabetes detection a study with nordic islet transplant and PIMA datasets

Diabetes Mellitus (DM) is a global health challenge, and accurate early detection is critical for effective management. The study explores the potential of machine learning for improved diabetes prediction using microarray gene expression data and PIMA data set. Researchers utilizing a hybrid feature extraction method such as Artificial Bee Colony (ABC) and Particle Swarm Optimization (PSO) followed by metaheuristic feature selection algorithms as Harmonic Search (HS), Dragonfly Algorithm (DFA), Elephant Herding Algorithm (EHA). Evaluated the performance of a system by using the following classifiers as Non-Linear Regression—NLR, Linear Regression—LR, Gaussian Mixture Model—GMM, Expectation Maximization—EM, Bayesian Linear Discriminant Analysis—BLDA, Softmax Discriminant Classifier—SDC, and Support Vector Machine with Radial Basis Function kernel—SVM-RBF classifier on two publicly available datasets namely the Nordic Islet Transplant Program (NITP) and the PIMA Indian Diabetes Dataset (PIDD). The findings demonstrate significant improvement in classification accuracy compared to using all genes. On the Nordic islet transplant dataset, the combined ABC-PSO feature extraction with EHO feature selection achieved the highest accuracy of 97.14%, surpassing the 94.28% accuracy obtained with ABC alone and EHO selection. Similarly, on the PIMA Indian diabetes dataset, the ABC-PSO and EHO combination achieved the best accuracy of 98.13%, exceeding the 95.45% accuracy with ABC and DFA selection. These results highlight the effectiveness of our proposed approach in identifying the most informative features for accurate diabetes prediction. It is observed that the parametric values attained for the datasets are almost similar. Therefore, this research indicates the robustness of the FE and FS along with classifier techniques with two different datasets.

Breast cancer classification based on hybrid CNN with LSTM model

Breast cancer (BC) is a global problem, largely due to a shortage of knowledge and early detection. The speed-up process of detection and classification is crucial for effective cancer treatment. Medical image analysis methods and computer-aided diagnosis can enhance this process, providing training and assistance to less experienced clinicians. Deep Learning (DL) models play a great role in accurately detecting and classifying cancer in the huge dataset, especially when dealing with large medical images. This paper presents a novel hybrid model of DL models combined a Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) for binary breast cancer classification on two datasets available at the Kaggle repository. CNNs extract mammographic features, including spatial hierarchies and malignancy patterns, whereas LSTM networks characterize sequential dependencies and temporal interactions. Our method combines these structures to improve classification accuracy and resilience. We compared the proposed model with other DL models, such as CNN, LSTM, Gated Recurrent Units (GRUs), VGG-16, and RESNET-50. The CNN-LSTM model achieved superior performance with accuracies of 99.17% and 99.90% on the respective datasets. This paper uses prediction evaluation metrics such as accuracy, sensitivity, specificity, F-score, and the AUC curve. The results showed that our model CNN-LSTM can enhance the performance of breast cancer classifiers compared with others with 99.90% accuracy on the second dataset.

CONVOLUTIONAL NEURAL NETWORK-BASED APPROACH FOR CLASSIFYING FUSARIUM WILT DISEASE IN CHICKPEAS USING IMAGE ANALYSIS

Legume crops, particularly chickpeas, are highly nutritious and play a vital role in global food security. However, they are susceptible to various diseases, among which Fusarium wilt, caused by Fusarium oxysporum, leads to significant yield losses. Early detection of Fusarium wilt is essential for effective disease management. Traditional diagnostic methods are often labour-intensive and time-consuming. This study aims to classify Fusarium wilt in chickpeas using Deep Convolutional Neural Networks (DCNN). The dataset consists of 4,339 chickpea plant images obtained from Kaggle. The images are categorized into five classes based on disease severity: highly resistant (HR), resistant (R), moderately resistant (MR), susceptible (S), and highly susceptible (HS). The images were pre-processed, resized, normalized, and augmented to enhance model performance. The classification was performed using a SoftMax classifier. The DCNN was trained using the Adam optimizer and categorical cross-entropy as the loss function, with hyperparameters fine-tuned to optimize performance. The proposed model achieved an overall accuracy of 73.96%, with a training accuracy of 73.16% and a validation accuracy of 77.64% after 100 epochs. Performance metrics revealed the highest precision and recall for the highly susceptible (HS) class, while accuracy was lower for intermediate classes (R and MR). The confusion matrix highlighted areas where the model excelled and where further refinement is needed. The study demonstrates the potential of DCNNs for automated classification of Fusarium wilt in chickpeas, offering a practical tool for disease management. However, the model's limitations in intermediate classes underline the need for further improvements. Future work will focus on enhancing dataset diversity, refining preprocessing techniques, and exploring advanced architectures to improve classification accuracy across all severity levels. These findings contribute to the development of robust, automated solutions for managing plant diseases and supporting sustainable agriculture. Keywords: Fusarium wilt, Chickpea, Deep Convolutional Neural Network (DCNN), Accuracy