Decoding how the brain computes with variables, such as an object's speed, encoded in populations of neurons is essential for understanding perception, decision-making, and motor control. While experiments consistently reveal correlated firing rates within these populations, the origins of these correlations are still debated. Here, I propose a computational model that carries out operations on population-encoded variables by carefully configuring the connections between neurons through dendritic branches. This model predicts neural correlations as a byproduct of the computation, with correlation coefficients aligning with experimentally measured values in the literature. I also present a mathematical derivation of these coefficients, revealing their dependence on population size. To demonstrate the model's utility, I apply it to compute the prediction error between an observed object's speed and the predicted speed based on self-motion. Importantly, the framework generalizes to compute any functional relationship between population-encoded variables, serving as a building block for perception and behaviour, including tasks such as transforming an object's position into joint angles for grasping.
In recent times, deep learning has been widely used in agriculture fields to identify diseases in crops, weather prediction, and crop yield prediction. However, designing efficient deep learning models that are lightweight, cost-effective, and suitable for deployment on small devices remains a challenge. This paper addresses this gap by proposing a Convolutional Neural Network (CNN) architecture optimized using a Genetic Algorithm (GA) to automate the selection of critical hyperparameters, such as the number and size of filters, ensuring high performance with minimal computational overhead. In this work, we have built our own tea leaf disease dataset consisting of three different tea leaf diseases, two diseases caused by pests, and one due to pathogens (infectious organisms) and environmental conditions. The proposed genetic algorithm-based CNN achieved an accuracy rate of 97.6% on the tea leaf disease dataset. To further validate its robustness, the model was tested on two additional datasets, namely PlantVillage and Rice leaf disease dataset, achieving accuracies of 96.99% and 99%, respectively. Performances of the proposed model are also compared with several state-of-the-art deep learning models, and the results show that the proposed model outperforms several DL architectures with fewer parameters.
Parkinson's Disease (PD) is a progressive neurodegenerative disorder and the early diagnosis is crucial for managing symptoms and slowing disease progression. This paper proposes a framework named Federated Learning Enabled Waterwheel Shuffled Shepherd Optimization-based Efficient-Fuzzy Deep Maxout Network (FedL_WSSO based Eff-FDMNet) for PD detection and classification. In local training model, the input image from the database "Image and Data Archive (IDA)" is given for preprocessing that is performed using Gaussian filter. Consequently, image augmentation takes place and feature extraction is conducted. These processes are executed for every input image. Therefore, the collected outputs of images are used for PD detection using Shepard Convolutional Neural Network Fuzzy Zeiler and Fergus Net (ShCNN-Fuzzy-ZFNet). Then, PD classification is accomplished using Eff-FDMNet, which is trained using WSSO. At last, based on CAViaR, local updation and aggregation are changed in server. The developed method obtained highest accuracy as 0.927, mean average precision as 0.905, lowest false positive rate (FPR) as 0.082, loss as 0.073, Mean Squared Error (MSE) as 0.213, and Root Mean Squared Error (RMSE) as 0.461. The high accuracy and low error rates indicate that the potent framework can enhance patient outcomes by enabling more reliable and personalized diagnosis.
Non-Orthogonal Multiple Access (NOMA) is the successive multiple-access methodologies for modern communication devices. Energy Efficiency (EE) is suggested in the NOMA system. In dynamic network conditions, the consideration of NOMA shows high computational complexity that minimizes the EE to degrade the system performance. This research suggested EE for the Multi-Carrier NOMA (MC-NOMA) models by optimization algorithm. The main scope of this research tends to improve the EE by Hybrid of Sewing Training and Lemur Optimization for optimizing the system parameters. The improvement made in this developed HSTLO algorithm can provide significant impact on MC-NOMA system, which it renders better user capacity while effectively optimizing the system parameters. Moreover, the Dilated Dense Recurrent Neural Network (DDRNN) model is developed. Employing the improvement in the deep learning model for the MC-NOMA system could effectively manage and enhance the system performance. Considering the DDRNN model can leverage to provide better generalization outcomes in different network scenarios that ensures to provide fast and reliable solutions compared to existing methods. Addressing the energy consumption problems in this research study will be analysed to show the advancement in MC-NOMA system that help to enhance the system performance.
Generally, electrically conductive materials are extremely sturdy and stiff, electric discharge milling (EDM) is a broadly utilized method. The usage of diamond grinding together with EDM in a machine is called the " and Electrical Discharge Diamond Grinding " (EDDG) gadget is an extensively used method for producing strong, long-lasting electrically conductive substances. The Modified Ant Lion Optimization- Artificial Neural Network (MALO-ANN) technique is recommended to boost the performance of EDDG machine. The MALO technique improves the overall performance of ANN by optimizing hidden layers and weights, which are regularly the cause of issues in traditional models. Input factors, along with grit size, pulse-on/off duration, height modern and pulse-off duration, are analysed to see if they affect Material Removal Rate (MRR) along with Surface Roughness (SR). The findings suggest that the MALO-ANN method greatly enhances the parametric optimization of EDDG gadget. The result indicates tremendous ability in improving the efficiency of EDDG systems, because conventional ANN models regularly struggle because of insifficient hidden layers and weights. The best MRR and SR were obtained with an absolute error interval ranging from 1.03% to 4.49%, achieving a convergence rate of 89%, performing enhanced accuracy in EDDG processes.
Existing Lung Cancer Diagnosis (LCD) models have difficulty in detecting early-stage lung cancer due to the asymptomatic nature of the disease which leads to an increased death rate of patients. Therefore, it is important to diagnose lung disease at an early stage to save the lives of affected persons. Hence, the research work aims to develop an efficient lung disease diagnosis using deep learning techniques for the early and accurate detection of lung cancer. This is achieved by. Initially, the proposed model collects the mandatory CT images from the standard benchmark datasets. Then, the lung cancer segmentation is done by using the development of Hybrid Convolution (2D/3D)-based Adaptive DenseUnet with Attention mechanism (HC-ADAM). The Hybrid Sewing Training with Spider Monkey Optimization (HSTSMO) is introduced to optimize the parameters in the developed HC-ADAM segmentation approach. Finally, the dissected lung nodule imagery is considered for the lung cancer classification stage, where the Hybrid Adaptive Dilated Networks with Attention mechanism (HADN-AM) are implemented with the serial cascading of ResNet and Long Short Term Memory (LSTM) for attaining better categorization performance. The accuracy, precision, and F1-score of the developed model for the LIDC-IDRI dataset are 96.3%, 96.38%, and 96.36%, respectively.
Speech enhancement techniques face computational demands, well-developed datasets, and better quality speech signals. Deep learners help deal with different noise types; still, the challenges offered by environmental noises require highly efficient and robust systems. This paper presents a lightweight deep-learning design with a heuristic-inspired model for generating an enhanced speech signal from noisy speech data. The model aims to remove different environmental noises affecting the speech signal. The noisy speech data are converted into spectrograms with Short-Time Fourier Transform (STFT). The noisy spectrogram is processed through the newly developed speech enhancement model namely, Layer Modified Residual Unet++ (LMResUnet++). The developed LMResUnet++ is designed through an atrous convolution layer, and it can capture multi-scale information without additional training parameter requirements. Also, the design is made compactable through the proposed hybrid optimization algorithm namely, Aquila Black Widow Optimization (ABWO), and it optimizes various hyperparameters of the developed LMResUnet++. The final denoised spectrogram from the LMResUnet++ undergoes Inverse STFT, and the final enhanced speech signal is restored. Further, different experiments are held to prove the efficacy of the system. Results prove that the developed LMResUnet++ achieved PESQ values of 7.93%, 5.75%, 3.86%, and 1.90% improved than DeepUnet, MTCNN, STCNN, and ResUnet++ respectively.
Image reconstruction is a critical step in various applications, such as art restoration, medical image processing, and agriculture, but it faces challenges due to noise and mosaic artefacts. In this research, a novel approach is introduced for de-noising and de-mosaicking images to enhance image reconstruction quality. The proposed model consists of three main steps: detail layer extraction, image de-noising using an Efficient Generative Adversarial Network (E-GAN), and de-mosaicking using an Adaptive Gannet-based Residual DenseNet (AG_DenseResNet). The publicly available Kodak dataset is utilized for the evaluation of the proposed model. The results show that the proposed outperforms conventional methods in terms of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), Mean Squared Error (MSE), and Learned Perceptual Image Patch Similarity (LPIPS) and acquired the values of 53.93, 0.98, 2.76, and 0.23, respectively.
This study explores the impacts of filter sizes on convolutional neural networks (CNNs) models, moving away from the common belief that small filters (3x3) give better results. The goal is to highlight the potential of large filters and encourage researchers to investigate their capabilities. The usage of large filters will increase the computational power which leads common researchers to reduce the filter size to reserve this power; however, other researchers address the potential of large filters to enhance the performance of CNN models. Currently, there are few pure CNN models that achieve optimal performance with large filters showing how far the large filter sizes topic is not addressed well by the community. As the availability of computer power and image sizes increase, traditional obstacles that hinder researchers from using large filter sizes will gradually diminish. This paper guides researchers by analysing and exploring the limitations, challenges, and impacts of CNN filter sizes on different CNN architectures. This will help utilize large filters' distinctive opportunities and potential. To our knowledge, we find four opportunities from utilizing large filters. A comprehensive comparison of researches on different CNN architectures shows a bias for small filters (3x3) and the possible potential of large filters.
In software development, software testing is very crucial for developing good quality software, where the effectiveness of software is to be tested. For software testing, test suites and test cases need to be prepared in minimum execution time with the test case prioritization (TCP) problems. Generally, some of the researchers mainly focus on the constraint problems, such as time and fault on TCP. In this research, the novel Fractional Hybrid Leader Based Optimization (FHLO) is introduced with constraint handling for combinatorial TCP. To detect faults earlier, the TCP is an important technique as it reduces the regression testing cost and prioritizes the test case execution. Based on the detected fault and branch coverage, the priority of the test case for program execution is decided. Furthermore, the FHLO algorithm establishes the TCP for detecting the program fault, which prioritizes the test case, and relies on maximum values of Average Percentage of Branch Coverage (APBC) and Average Percentage of Fault Detected (APFD). From the analysis, the devised FHLO algorithm attains a maximum value of 0.966 for APFD and 0.888 for APBC.
The anomalous enlargement of brain cells is known as Brain Tumour (BT), which can cause serious damage to different blood vessel and nerve in human body. A precise and early detection of BT is foremost important to eliminate severe illness. Thus, a SpinalNet Visual Geometry Group-16 (Spinal VGG-16-Net) is introduced for early BT detection. At first, Magnetic Resonance Imaging (MRI) of image obtained from data sample is subjected to image denoising by bilateral filter. Then, BT area is segmented from the image using entropy-based Kapur thresholding technique, where threshold values are ideally selected by Gradient Energy Valley Optimization (GEVO), which is designed by incorporating Energy Valley Optimization (EVO) with Stochastic Gradient Descent (SGD) algorithm. Then, process of image augmentation is worked and later, feature extraction is performed to mine the most significant features. Finally, BT is detected using proposed Spinal VGG-16Net, which is devised by combining both SpinalNet and VGG-16 Net. The Spinal VGG-16-Net is compared with some of the existing schemes, and it attained maximum accuracy of 92.14%, True Positive Rate (TPR) of 93.16%, True Negative Rate (TNR) of 91.35%, Negative Predictive Value (NPV) 89.73%, and Positive Predictive Value (PPV) o of 92.13%.
With the development of medical imaging amenities, a rising quantity of data emerges in the present image processing that has led to gradually more burden for data transmission and storage. Image compression is a method of lessening the excess in images and symbolizing it in a short way that could permit more gainful exploitation of storage capacity and network bandwidth. This paper develops a new image compression model with steps like segmentation, encoding, and decoding. Initially, segmentation is carried out using Improved Hierarchical Density-Based Spatial Clustering of Applications with Noise (HDBSCAN). This phase assists in ROI separation. Subsequently, compression occurs using Improved Huffman encoding. Also, in particular, the encoding parameters are optimally chosen via a new algorithm named Snake Updated BES Optimization (SU-BESO). In the last phase, decoding is done, during which, Huffman decoding as well as region fusion are carried out. Finally, the examination is done to prove the potential of the developed SU-BESO model.
PURPOSE:Big Data's extensive capabilities can aid in addressing the unpredictability of food supply caused by a variety of issues including soil degradation, climate change, water pollution, socio-cultural expansion, governmental laws, and market volatility. However, crop monitoring and classification are critical components of agricultural precision farming. This paper intends to propose a crop classification via a hybrid classification model. DESIGN:First, the input image dataset is subjected to the preprocessing stage to enhance the image dataset by removing noise and blurring the edges with the aid of Gaussian filtering. Second, the improved spider local image feature, median binary pattern and haralick texture features are extracted from the preprocessed image dataset by utilizing the map-reduce framework, to handle big data. Third, the hybrid classification model is proposed that involves two classifiers such as Bi-GRU and CNN. FINDINGS:The weights of both classifier Bi-GRU and CNN were tuned optimally by the proposed hybrid optimization BUBMO that combined both BMO and BWO. The greatest MCC obtained by the propose is 91.47%, whilst the traditional model scored the lowest MCC. ORIGINALITY:The accuracy and improved efficacy of the crop categorization are achieved by employing the suggested classification method.
Brain tumour classification plays a significant role in improving patient care, treatment planning, and enhancing the overall healthcare system's effectiveness. This article presents a ResNet framework optimized using Henry gas solubility optimization (HGSO) for the classification of brain tumours, resulting in improved classification performance in magnetic resonance images (MRI). Two variants of the deep residual neural network, namely ResNet-18 and ResNet-50, are trained on the MRI training dataset. The four critical hyperparameters of the ResNet model: momentum, initial learning rate, maximum epochs, and validation frequency are tuned to obtain optimal values using HGSO algorithm. Subsequently, the optimized ResNet model is evaluated using two separate databases: Database1, comprising four tumour classes, and Database2, with three tumour classes. The performance is assessed using accuracy, sensitivity, specificity, precision, and F-score. The highest classification accuracy of 0.9825 is attained using the proposed optimized ResNet-50 framework on Database1. Moreover, the Gradient-weighted Class Activation Mapping (GRAD-CAM) algorithm is utilized to enhance the understanding of deep neural networks by highlighting the regions that are influential in making a particular classification decision. Grad-CAM heatmaps confirm the model focuses on relevant tumour features, not image artefacts. This research enhances MRI brain tumour classification via deep learning optimization strategies.
The sentiment analysis is an essential component that enables automation of achieving insights from the information that is user generated. However, the difficulty of sentiment analysis is the lack of enough labelled data in the Natural Language Processing (NLP) sector. Thus, to evaluate these sentiments, multiple mechanisms have been utilized in the past decades. The deep learning-aided approaches are becoming very famous nowadays because of their better performances. To surmount such existing issues, an attention deep learning model is proposed using an improved heuristic approach. At first, the input text data is gathered from public resources. Further, it is followed by text pre-processing to prevent unrelated text data. Further, the obtained pre-processed text is fed into the Multiscale Feature Fusion-based Adaptive and Attention-based Convolution Neural Network (MFF-AACNet). In the developed system, the features are extracted from Bidirectional Encoder Representations from Transformers (BERT), Transformers, and word2vector. Furthermore, the resultant features are fused, and it is subjected to the MFF-AACNet, where the sentiment is analysed. The parameter tuning is done by an improved Fitness Opposition of Rat Swarm Optimizer (FORSO). Finally, the performance analysis was conducted for the implemented model. The proposed framework achieves higher accuracy compared to traditional methods.
In this study, a novel technique termed Quadratic Discriminant Feature Selected Broken Stick Regressive Deep Convolution Neural Learning Classification (QDFSBSRDCNLC) Technique is proposed for disease classification and hence yields prediction of turmeric crop. Initially, we gathered the images of turmeric crops with and without diseases. The images are collected from the turmeric research field at Bhavanisagar. Quadratic Discriminant Analysis (QDA) is utilized to select relevant features from a dataset, reducing dimensionality. In this paper, four models, named FCN8, PSP Net, MobileNetV3 (small), and Deep Lab V3 are chosen for semantic segmentation of disease in turmeric crops. Turmeric crop production predicts is an important part of modern agriculture, allowing farmers to make sensible choices and optimize resources. We can predict turmeric crop yields accurately by using modern data analysis approaches. Predictive models take into consideration variables such as weather, soil quality, and farming techniques. The experimental results demonstrated that MobileNetV3 (small) performed better than other established ones with the accuracy of 97.99%, IoU of 96.82%, and Coefficient of 97.80% for 50 epochs. The proposed QDFSBSRDCNLC Technique effectively classifies diseases and predicts the yield of turmeric crops, with MobileNetV3 (small) showing superior performance among the tested models.
In this paper, meningioma detection and segmentation method is proposed. This research work proposes an effective method to locate meningioma pictures through a novel CFCM classification approach. This proposed method consist of Non-Sub sampled Contourlet Transform decomposition module which decomposes the entire brain image into multi-scale sub-band images and then the heuristic and uniqueness features have been computed individually. Then, these heuristic and uniqueness features are trained and classified using Convolutional Fuzzy C Means (CFCM) classifier. This proposed method is applied on two independent brain imaging datasets. The proposed meningioma identification system stated in this work obtained 98.81% of Se, 98.83% of Sp, 99.04% of Acc, 99.12% of pr, and 99.14% of FIS on Nanfang University dataset brain images. The proposed meningioma identification system stated in this work obtained 98.92% of Se, 98.88% of Sp, 98.9% of Acc, 98.88% of pr, and 99.36% of FIS on the BRATS 2021 brain images. Finally, the tumour segmentation module is designed in VLSI, and it is simulated using Xilinx project navigator in this paper.
The existing approaches suffer from scalability and security issues while transmitting data. Blockchain is a recently emerged technology, and it is an emerging platform that allows secure transmission. A distributed design is required to address these issues and abide by security regulations. Blockchain has been recently introduced as an alternative solution to solve complex and challenging security issues while storing data. Thus, an intelligent blockchain-assisted IoT architecture is provided in this work to perform secure healthcare data transmission. The first aim of our model is to detect malware attacks in IoT networks. To detect the malware activities, the attack detection data was gathered, and it was fed as input to the Hybrid Adaptive Deep Learning Method. For further enhancement, the FUPOA performs the parameter tuning. A privacy preservation model is employed to secure healthcare data by generating the optimal key formation, in which the key is optimized using FUPOA. This secured data can be stored in the blockchain to increase data integrity and privacy. The optimal feature selection is done by the FUPOA approach. Further, the acquired optimal features are fed to the HADL-AM for predicting the data. The experimental analysis has been done and compared among different approaches.
Brain Tumor (BT) is an irregular growth of cells in the brain or in the tissues surrounding it. Detecting and predicting tumours is essential in today's world, yet managing these diseases poses a considerable challenge. Among the various modalities, Magnetic Resonance Imaging (MRI) has been extensively exploited for diagnosing tumours. The traditional methods for predicting survival are based on handcrafted features from MRI and clinical information, which is generally subjective and laborious. This paper devises a new method named, Deep Residual PyramidNet (DRP_Net) for BT classification and survival prediction. The input MRI image is primarily derived from the BraTS dataset. Then, image enhancement is done to improve the quality of images using homomorphic filtering. Next, deep joint segmentation is used to process the tumourtumour region segmentation. Consequently, Haar wavelet and Local Directional Number Pattern (LDNP) based feature extraction is mined. Afterward, BT classification is achieved through DRP_Net, which is a fusion of Deep Residual Network (DRN) and PyramidNet. At last, the survival prediction is accomplished by employing the Deep Recurrent Neural Network (DRNN). Furthermore, DRP_Net has attained superior performance with a True Negative Rate (TNR) of 91.99%, an accuracy of 90.18%, and True Positive Rate (TPR) of 91.08%.
Efficient data transmission in Wireless Sensor Networks (WSNs) is a critical challenge. Traditional routing protocols focus on energy efficiency but do not consider other factors that might degrade performance. This research proposes a novel Hybrid Beluga Whale-Coati Optimization (HBWCO) algorithm to address these issues, focusing on optimizing energy-efficient data transmission. In the proposed approach, initially, sensor nodes and field dimensions are initialized. Then, K-means clustering is applied to grouping nodes. The Deep Q-Net model is used to predict energy levels of nodes. CH is selected as per the node having higher energy. Multipath routing is performed through the HBWCO algorithm, which optimally selects the best routing paths by considering factors like reliability, residual energy, predicted energy, throughput, and traffic intensity. If link breakage occurs, a route maintenance phase is initiated using Source Link Breakage Warning (SLBW) message strategy to notify the source node about the issue of choosing another path. This work offers a comprehensive approach to enhancing energy efficiency in networks. The suggested HBWCO approach is in contrast to the traditional methods. The HBWCO approach has achieved the highest reliability of 0.948 and the highest throughput of 3496. Therefore, the HBWCO algorithm offers an effective solution for data transmission and routing reliability.
