Performance Of Deep Learning Algorithms Research Articles

A Mountain Gazelle Optimization (MGO) for Enhancing the Deep Learning Performance in Various Operating Systems

This study introduces a novel optimization framework that assesses and enhances deep learning algorithm performance across autonomous car operating systems. The framework generates synthetic performance measures, including inference time, memory use, CPU/GPU utilization, and accuracy, to evaluate algorithm performance. By employing the Mountain Gazelle Optimizer (MGO), the study identifies the best deep learning algorithm and operating system setup for accuracy and resource efficiency. The proposed methodology normalizes performance indicators, defines a fitness function to assist optimization, and then iterates through numerous configurations to discover the optimum option. Extensive trials and scenario comparisons validate the effectiveness of our approach. Computational efficiency and accuracy improved significantly, revealing the ideal autonomous car system performance combinations. This study not only enhances deep learning optimization but also provides practical instructions for building robust autonomous car systems in varied operating contexts, thereby informing the development of future autonomous vehicles and offering actionable insights for developers and researchers in the field.

Profile Photograph Classification Performance of Deep Learning Algorithms Trained Using Cephalometric Measurements: A Preliminary Study.

Extraoral profile photographs are crucial for orthodontic diagnosis, documentation, and treatment planning. The purpose of this study was to evaluate classifications made on extraoral patient photographs by deep learning algorithms trained using grouped patient pictures based on cephalometric measurements. Cephalometric radiographs and profile photographs of 990 patients from the archives of Kocaeli University Faculty of Dentistry Department of Orthodontics were used for the study. FH-NA, FH-NPog, FMA and N-A-Pog measurements on patient cephalometric radiographs were carried out utilizing Webceph. 3 groups for every parameter were formed according to cephalometric values. Deep learning algorithms were trained using extraoral photographs of the patients which were grouped according to respective cephalometric measurements. 14 deep learning models were trained and tested for accuracy of prediction in classifying patient images. Accuracy rates of up to 96.67% for FH-NA groups, 97.33% for FH-NPog groups, 97.67% for FMA groups and 97.00% for N-A-Pog groups were obtained. This is a pioneering study where an attempt was made to classify clinical photographs using artificial intelligence architectures that were trained according to actual cephalometric values, thus eliminating or reducing the need for cephalometric X-rays in future applications for orthodontic diagnosis.

Performance of a vectorcardiographic deep learning algorithm compared to single-lead and 12-lead ECG for atrial flutter detection: implications for wearable devices

Abstract Background Wearable devices have developed the capacity to offer single-lead ECG analysis and are already widespread in use, particularly for detection of atrial fibrillation (1,2). Vectorcardiography entails the transformation of electrical vectors into X, Y, and Z axes to constitute a vectorcardiogram (VCG) (3). The VCG has diagnostic promise given the capacity to generate it without requiring clinical presentation for standard 12-lead ECG acquisition. Furthermore, a novel credit-card sized device capable of acquiring a VCG has recently been developed – raising the potential for arrhythmia detection and discrimination beyond wearable single-lead ECG systems. Objective To assess the performance of a deep learning algorithm trained on single lead and 12-lead ECGs and VCGs for detecting atrial flutter. Methods A convolutional neural network (AFL-CNN) was developed on a dataset of 4,000 ECGs and VCGs from a single US academic medical institution (80% training and 20% for testing). Validation was performed on a separate publicly available dataset (2223 ECGs). Model performance was assessed on single lead ECG analysis (lead I, obtained by removing the other leads) and standard 12-lead ECG. The gold standard for AFL diagnosis was physician-confirmed 12-lead ECG diagnoses in the dataset. The VCG was generated from the 12-lead ECG via Kors transformation matrix to which AFL-CNN was also applied. Model performance was assessed with sensitivity, specificity, positive and negative predictive values, accuracy, and F1 statistic. Results On single lead ECG analysis, AFL-CNN achieved 71.2% sensitivity and 96.9% specificity with PPV 84.0% and NPV 93.7% and 92.2% accuracy with F1 83.2%. When AFL-CNN was applied to the VCG, performance improved: 91.0% sensitivity and 98.7% specificity, PPV 94.24%, NPV 98.0%, 97.3% accuracy, and F1 95.31%. Lastly, the performance remained high when AFL-CNN was applied to the 12-lead ECG: 95.4% sensitivity, 98.63% specificity, PPV 94.0%, NPV 99.0%, and accuracy of 98.0%, with F1 97.2%. Conclusion A deep learning algorithm applied to vectorcardiography for the detection of atrial flutter showed improved performance over single lead analysis and comparable performance to standard 12-lead applications. The finding of improved model performance beyond single lead ECG tracings underscores both the limitations of current wearable technologies in arrhythmia detection beyond atrial fibrillation and the potential for VCG-based algorithm prediction to fill gaps in healthcare inequality when standard 12-lead acquisition is challenging to obtain.

Deep Learning Assessment of Small Renal Masses at Contrast-enhanced Multiphase CT.

Background Accurate characterization of suspicious small renal masses is crucial for optimized management. Deep learning (DL) algorithms may assist with this effort. Purpose To develop and validate a DL algorithm for identifying benign small renal masses at contrast-enhanced multiphase CT. Materials and Methods Surgically resected renal masses measuring 3 cm or less in diameter at contrast-enhanced CT were included. The DL algorithm was developed by using retrospective data from one hospital between 2009 and 2021, with patients randomly allocated in a training and internal test set ratio of 8:2. Between 2013 and 2021, external testing was performed on data from five independent hospitals. A prospective test set was obtained between 2021 and 2022 from one hospital. Algorithm performance was evaluated by using the area under the receiver operating characteristic curve (AUC) and compared with the results of seven clinicians using the DeLong test. Results A total of 1703 patients (mean age, 56 years ± 12 [SD]; 619 female) with a single renal mass per patient were evaluated. The retrospective data set included 1063 lesions (874 in training set, 189 internal test set); the multicenter external test set included 537 lesions (12.3%, 66 benign) with 89 subcentimeter (≤1 cm) lesions (16.6%); and the prospective test set included 103 lesions (13.6%, 14 benign) with 20 (19.4%) subcentimeter lesions. The DL algorithm performance was comparable with that of urological radiologists: for the external test set, AUC was 0.80 (95% CI: 0.75, 0.85) versus 0.84 (95% CI: 0.78, 0.88) (P = .61); for the prospective test set, AUC was 0.87 (95% CI: 0.79, 0.93) versus 0.92 (95% CI: 0.86, 0.96) (P = .70). For subcentimeter lesions in the external test set, the algorithm and urological radiologists had similar AUC of 0.74 (95% CI: 0.63, 0.83) and 0.81 (95% CI: 0.68, 0.92) (P = .78), respectively. Conclusion The multiphase CT-based DL algorithm showed comparable performance with that of radiologists for identifying benign small renal masses, including lesions of 1 cm or less. Published under a CC BY 4.0 license. Supplemental material is available for this article.

DWT-BILSTM-based models for day-ahead hourly global horizontal solar irradiance forecasting

Accurate forecasting of electricity generation from renewable energy sources is crucial for the operation, planning and management of smart grids. For reliable planning and operation of photovoltaic (PV) systems in grid-connected or islanded utilities, an hourly day-ahead forecast of PV output is critical. The forecast of PV power can be done indirectly by estimating solar irradiance. For forecasting day-ahead hourly global horizontal irradiance (GHI), two forecasting models with different multivariate inputs are proposed in this paper, and the results are compared. These models use a hybrid algorithm of discrete wavelet decomposition and bidirectional long short-term memory (BILSTM). The inputs of the first model contain GHI and weather type data. The other model allows for observation of the effect of meteorological values including GHI, temperature, humidity, wind speed, and weather type data. The forecasting performance of deep learning algorithms which contain recurrent neural network (RNN), long short-term memory (LSTM), and BILSTM algorithms for day ahead hourly solar irradiance forecasting problems are also compared. To evaluate the performance of proposed models, two datasets are used for Model 1 and one dataset is used for Model 2. An experiment is also done to demonstrate that the proposed Model 1 is applicable in datasets collected in the vicinity of the city of Trabzon. On the other hand, BILSTM algorithm outperforms RNN and LSTM algorithms. It is seen that the test successes of both proposed models are better than the results given in the literature.

A time-delay neural network for ship-radiated noise recognition based on residual block and attention mechanism

Ship-radiated noise recognition remains a challenging issue in underwater acoustic signal processing for a variety of reasons. The ocean ambient noise influences the signal quality, and the scarcity of samples limits the performance of deep learning algorithms. In this study, a self-attention mechanism and residual block-based time-delay neural network (SRTDNN) is proposed to improve the recognition performance of ship-radiated noise. The noise samples are first segmented and transformed into three-dimensional Mel-scale frequency cepstral coefficients (3D-MFCC), which combine static and dynamic characteristics of targets through the concatenation of MFCC, Delta MFCC and Delta-Delta MFCC. Subsequently, 3D-MFCC are fed as input to the SRTDNN. The SRTDNN extracts time-domain features of inputs through the TDNN layers and then assigns different weights to the frame and channel dimensions of the feature maps through two attention mechanisms. The Res Connections and Res2Block are interspersed to facilitate convergence. During training, a time-domain and frequency-domain data augmentation pipeline is designed to alleviate the lack of samples. Experiments with public and real sea experiment datasets are conducted to compare the SRTDNN with acknowledged high-quality recognition methods. The results show that the SRTDNN has a significant improvement in recognition performance. Furthermore, the ablation tests demonstrate the effectiveness of the proposed method.

Advanced Soil Organic Matter Prediction with a Regional Soil NIR Spectral Library Using Long Short-Term Memory–Convolutional Neural Networks: A Case Study

Soil analysis using near-infrared spectroscopy has shown great potential to be an alternative to traditional laboratory analysis, and there is continuously increasing interest in building large-scale soil spectral libraries (SSLs). However, due to issues such as high non-linearity in soil spectral data and complexity in soil spatial variation, the establishment of robust prediction models for soil spectral libraries remains a challenge. This study aimed to investigate the performance of deep learning algorithms, including long short-term memory (LSTM) and LSTM–convolutional neural networks (LSTM–CNN) integrated models, to predict the soil organic matter (SOM) of a provincial-scale SSL, and compare it to the normally used local weighted regression (LWR) model. The Hebei soil spectral library (HSSL) contains 425 topsoil samples (0–20 cm), of which every 3 soil samples were collected from dry land, irrigated land, and paddy fields, respectively, in different counties of Hebei Province, China. The results show that the accuracy of the validation dataset rank as follows: LSTM–CNN (R2p = 0.96, RMSEp = 1.66 g/kg) > LSTM (R2p = 0.83, RMSEp = 3.42 g/kg) > LWR (R2p = 0.82, RMSEp = 3.79 g/kg). The LSTM–CNN model performed the best, mainly due to its comprehensive ability to effectively extract spatial and temporal features. Meanwhile, the LSTM model achieved higher accuracy than the LWR model, owing to its built-in memory unit and its advantage of faster feature band extraction. Thus, it was suggested to use deep learning algorithms for SOM predictions in SSLs. However, their performance on larger-scale SSLs such as continental/global SSLs still needs to be further investigated.

Assessing the Performance of Deep Learning Algorithms With and Without Transfer Learning in Similar or Identical Wastewater Treatment Processes

This study assessed the feasibility of transfer learning from one wastewater treatment process to another using two popular deep learning algorithms. Specifically, convolutional neural network (CNN) and long short-term memory (LSTM), which consisted of four and three hidden layers, respectively, were used as benchmark algorithms for transfer learning. Input data for both deep learning and transfer learning were provided from two wastewater treatment plants with identical treatment trains in series (located in Jinju and Cheongju City) over the five-year period from 2018 to 2022. Performance evaluation was also done not only against two deep learning algorithms but also against those adopting two transfer learning strategies, one for freezing all hidden layers developed from the pre-trained model and the other for training the last hidden layer only among multiple ones, with respect to Mean Squared Error (MSE). We found that the performance of both CNN and LSTM was relatively comparative regardless of dependent variables, discharge and biochemical oxygen demand (BOD), whereas the prediction accuracy of both algorithms was slightly higher for discharge than for BOD due to its low variability. When transfer learning which froze all hidden layers of the existing model was applied to two benchmark algorithms, the predictive performance of both algorithms was found to slightly improved only for discharge. Also, there was no measurable variation in the prediction accuracy of benchmark algorithms using the other transfer learning approach. Potential applications of transfer learning include the rapid reuse of the existing models (developed from source domains) for target domains which are hard to develop new prediction models due to the lack of data in deep learning.

Performance of anterior segment OCT-based algorithms in the opportunistic screening for primary angle-closure disease

PurposeThis study aimed to investigate the performance of deep learning algorithms in the opportunistic screening for primary angle-closure disease (PACD) using combined anterior segment parameters. MethodsThis was an observational, cross-sectional hospital-based study. Patients with PACD and healthy controls who underwent comprehensive eye examinations, including gonioscopy and anterior segment optical coherence tomography (ASOCT) examinations under both light and dark conditions, were consecutively enrolled from the Department of Ophthalmology at the Beijing Tongren Hospital between November 2020 and June 2022. The anterior chamber, anterior chamber angle, iris, and lens parameters were assessed using ASOCT. To build the prediction models, backward logistic regression was utilized to select the variables to discriminate patients with PACD from normal participants, and the area under the receiver operating characteristic curve was used to evaluate the efficacy of the opportunistic screening. ResultsThe data from 199 patients (199 eyes) were included in the final analysis and divided into two groups: PACD (109 eyes) and controls (90 eyes). Angle opening distance at 500 μm, anterior chamber area, and iris curvature measured in the light condition were included in the final prediction models. The area under the receiver operating characteristic curve was 0.968, with a sensitivity of 91.74 % and a specificity of 91.11 %. ConclusionASOCT-based algorithms showed excellent diagnostic performance in the opportunistic screening for PACD. These results provide a promising basis for future research on the development of an angle-closure probability scoring system for PACD screening.

A polar transformation augmentation approach for enhancing mammary gland segmentation in ultrasound images

Environmental factors can detrimentally affect mammary gland development, leading to negative impacts on milk secretion in mammals. Ultrasonography serves as a non-invasive and non-destructive method for assessing mammary gland characteristics and development. Deep learning approaches enable automated monitoring of mammary gland development, though they typically require large, labeled datasets that may be limited by data collection constraints. This study aimed to develop and evaluate a polar transformation-based augmentation strategy to enhance the performance of deep learning algorithms for mammary gland segmentation in small datasets. We collected 405 ultrasound images of mammary glands (front and rear quarters) from 29 crossbred F1 Holstein x Gyr calves aged 5 to 11 weeks. The parenchyma tissue in these images was manually annotated using the VGG Image Annotator. A leave-one-animal-out cross-validation approach was employed to train the semantic segmentation algorithm. In this approach, all images from one calf were used as a testing set, and images from the remaining 28 calves were used for training in each of the 29 iterations. Our proposed method involved utilizing a polar transform technique for data augmentation in ultrasound images and the PSPNet deep learning algorithm for image segmentation. The average F1-score on the testing set was 54% in week 1, 70% in week 2, and 75% in week 3. Our findings revealed that the algorithm’s performance was suboptimal for images with very small parenchyma (week 1). However, as the mammary gland developed, the identification and segmentation of parenchymal tissue significantly improved. The performance of deep learning algorithms in segmenting small tissues could potentially be enhanced by using larger datasets and higher resolution images. In conclusion, our study demonstrates that polar transformation is an effective strategy for augmenting mammary gland ultrasound images, which in turn improves the performance of deep neural networks in segmenting parenchymal tissue.

FORECASTING THE EURO EXCHANGE RATE USING DEEP LEARNING ALGORITHMS AND MACHINE LEARNING ALGORITHMS

Given that time series forecasts are of great importance in the financial world, the main objective of this study is to forecast Euro prices and examine the contribution of these forecasts to financial decision-making processes. Since the Euro is an important component of international trade and investment, accurate price forecasts are of strategic importance for many financial institutions and investors. In this study, we compare the performance of deep learning algorithms and classical machine learning methods for forecasting Euro prices: support vector machines (SVM), Extreme Gradient Boosting (XGBoost), long short-term memory (LSTM), and gated recurrent units (GRU). These methods represent different algorithms that are widely used in financial forecasting and give successful results. The dataset used in the study was divided into two parts: 80% training and 20% testing, and it is also indicated how each algorithm behaved during the training process and which parameters were chosen. The results are presented by comparing the performance of these algorithms, and it is found that the GRU algorithm provides better accuracy than the others. Therefore, the GRU algorithm was chosen to forecast Euro prices for the next 12 months, and the forecasting process was carried out. The results of this study are expected to provide an important perspective to financial decision-makers by comprehensively comparing the performance of deep learning and traditional approaches in Euro price forecasting. It also includes potential research avenues for future work and suggestions for the development of new methods in this area.

Deep learning-based modulation recognition with constellation diagram: A case study

Automatic modulation recognition is a promising solution for identifying and classifying signals received in heterogeneous wireless networks. In dynamic and autonomous environments, receivers must extract the relevant signal from various modulated signals to enable further communication procedures. Machine learning, including its sub-branches for classification problems, offers promising operational capabilities. This study utilized the ResNet-50 deep learning method for modulation classification. A dataset consisting of eight digital modulation techniques was generated, with constellation diagrams created as image data over the additive white Gaussian noise (AWGN) channel at signal-to-noise ratios (SNR) of 5 dB, 10 dB, and 20 dB. The deep learning algorithm's performance metrics were evaluated using a confusion matrix, and F1 scores were compared to those of the AlexNet deep learning algorithm. The simulation results clearly indicate the superior performance of ResNet-50 over AlexNet. In terms of average F1 scores, ResNet-50 exhibits a significant advantage, surpassing AlexNet by approximately 67%, 29%, and 10% at SNR values of 5 dB, 10 dB, and 20 dB, respectively.

A Review of Object Detection in Traffic Scenes Based on Deep Learning

Abstract At the current stage, the rapid Development of autonomous driving has made object detection in traffic scenarios a vital research task. Object detection is the most critical and challenging task in computer vision. Deep learning, with its powerful feature extraction capabilities, has found widespread applications in safety, military, and medical fields, and in recent years has expanded into the field of transportation, achieving significant breakthroughs. This survey is based on the theory of deep learning. It systematically summarizes the Development and current research status of object detection algorithms, and compare the characteristics, advantages and disadvantages of the two types of algorithms. With a focus on traffic signs, vehicle detection, and pedestrian detection, it summarizes the applications and research status of object detection in traffic scenarios, highlighting the strengths, limitations, and applicable scenarios of various methods. It introduces techniques for optimizing object detection algorithms, summarizes commonly used object detection datasets and traffic scene datasets, along with evaluation criteria, and performs comparative analysis of the performance of deep learning algorithms. Finally, it concludes the development trends of object detection algorithms in traffic scenarios, providing research directions for intelligent transportation and autonomous driving.

Image segmentation of impacted mesiodens using deep learning.

This study aimed to evaluate the performance of deep learning algorithms for the classification and segmentation of impacted mesiodens in pediatric panoramic radiographs. A total of 850 panoramic radiographs of pediatric patients (aged 3-9 years) was included in this study. The U-Net semantic segmentation algorithm was applied for the detection and segmentation of mesiodens in the upper anterior region. For enhancement of the algorithm, pre-trained ResNet models were applied to the encoding path. The segmentation performance of the algorithm was tested using the Jaccard index and Dice coefficient. The diagnostic accuracy, precision, recall, F1-score and time to diagnosis of the algorithms were compared with those of human expert groups using the test dataset. Cohen's kappa statistics were compared between the model and human groups. The segmentation model exhibited a high Jaccard index and Dice coefficient (>90%). In mesiodens diagnosis, the trained model achieved 91-92% accuracy and a 94-95% F1-score, which were comparable with human expert group results (96%). The diagnostic duration of the deep learning model was 7.5 seconds, which was significantly faster in mesiodens detection compared to human groups. The agreement between the deep learning model and human experts is moderate (Cohen's kappa = 0.767). The proposed deep learning algorithm showed good segmentation performance and approached the performance of human experts in the diagnosis of mesiodens, with a significantly faster diagnosis time.

The 3D-aware image synthesis of prohibited items in the X-ray security inspection by stylized generative radiance fields

<abstract><p>The merging of neural radiance fields with generative adversarial networks (GANs) can synthesize novel views of objects from latent code (noise). However, the challenge for generative neural radiance fields (NERFs) is that a single multiple layer perceptron (MLP) network represents a scene or object, and the shape and appearance of the generated object are unpredictable, owing to the randomness of latent code. In this paper, we propose a stylized generative radiance field (SGRF) to produce 3D-aware images with explicit control. To achieve this goal, we manipulated the input and output of the MLP in the model to entangle and disentangle label codes into/from the latent code, and incorporated an extra discriminator to differentiate between the class and color mode of the generated object. Based on the labels provided, the model could generate images of prohibited items varying in class, pose, scale, and color mode, thereby significantly increasing the quantity and diversity of images in the dataset. Through a systematic analysis of the results, the method was demonstrated to be effective in improving the detection performance of deep learning algorithms during security screening.</p></abstract>

Select, Purify, and Exchange: A Multisource Unsupervised Domain Adaptation Method for Building Extraction.

Accurately extracting buildings from aerial images has essential research significance for timely understanding human intervention on the land. The distribution discrepancies between diversified unlabeled remote sensing images (changes in imaging sensor, location, and environment) and labeled historical images significantly degrade the generalization performance of deep learning algorithms. Unsupervised domain adaptation (UDA) algorithms have recently been proposed to eliminate the distribution discrepancies without re-annotating training data for new domains. Nevertheless, due to the limited information provided by a single-source domain, single-source UDA (SSUDA) is not an optimal choice when multitemporal and multiregion remote sensing images are available. We propose a multisource UDA (MSUDA) framework SPENet for building extraction, aiming at selecting, purifying, and exchanging information from multisource domains to better adapt the model to the target domain. Specifically, the framework effectively utilizes richer knowledge by extracting target-relevant information from multiple-source domains, purifying target domain information with low-level features of buildings, and exchanging target domain information in an interactive learning manner. Extensive experiments and ablation studies constructed on 12 city datasets prove the effectiveness of our method against existing state-of-the-art methods, e.g., our method achieves 59.1% intersection over union (IoU) on Austin and Kitsap → Potsdam, which surpasses the target domain supervised method by 2.2% . The code is available at https://github.com/QZangXDU/SPENet.

Comparison of MRI radiomics-based machine learning survival models in predicting prognosis of glioblastoma multiforme

ObjectiveTo compare the performance of radiomics-based machine learning survival models in predicting the prognosis of glioblastoma multiforme (GBM) patients.Methods131 GBM patients were included in our study. The traditional Cox proportional-hazards (CoxPH) model and four machine learning models (SurvivalTree, Random survival forest (RSF), DeepSurv, DeepHit) were constructed, and the performance of the five models was evaluated using the C-index.ResultsAfter the screening, 1792 radiomics features were obtained. Seven radiomics features with the strongest relationship with prognosis were obtained following the application of the least absolute shrinkage and selection operator (LASSO) regression. The CoxPH model demonstrated that age (HR = 1.576, p = 0.037), Karnofsky performance status (KPS) score (HR = 1.890, p = 0.006), radiomics risk score (HR = 3.497, p = 0.001), and radiomics risk level (HR = 1.572, p = 0.043) were associated with poorer prognosis. The DeepSurv model performed the best among the five models, obtaining C-index of 0.882 and 0.732 for the training and test set, respectively. The performances of the other four models were lower: CoxPH (0.663 training set / 0.635 test set), SurvivalTree (0.702/0.655), RSF (0.735/0.667), DeepHit (0.608/0.560).ConclusionThis study confirmed the superior performance of deep learning algorithms based on radiomics relative to the traditional method in predicting the overall survival of GBM patients; specifically, the DeepSurv model showed the best predictive ability.

A new framework for water quality forecasting coupling causal inference, time-frequency analysis and uncertainty quantification

Accurate forecasting of water quality variables in river systems is crucial for relevant administrators to identify potential water quality degradation issues and take countermeasures promptly. However, pure data-driven forecasting models are often insufficient to deal with the highly varying periodicity of water quality in today's more complex environment. This study presents a new holistic framework for time-series forecasting of water quality parameters by combining advanced deep learning algorithms (i.e., Long Short-Term Memory (LSTM) and Informer) with causal inference, time-frequency analysis, and uncertainty quantification. The framework was demonstrated for total nitrogen (TN) forecasting in the largest artificial lakes in Asia (i.e., the Danjiangkou Reservoir, China) with six-year monitoring data from January 2017 to June 2022. The results showed that the pre-processing techniques based on causal inference and wavelet decomposition can significantly improve the performance of deep learning algorithms. Compared to the individual LSTM and Informer models, wavelet-coupled approaches diminished well the apparent forecasting errors of TN concentrations, with 24.39%, 32.68%, and 41.26% reduction at most in the average, standard deviation, and maximum values of the errors, respectively. In addition, a post-processing algorithm based on the Copula function and Bayesian theory was designed to quantify the uncertainty of predictions. With the help of this algorithm, each deterministic prediction of our model can correspond to a range of possible outputs. The 95% forecast confidence interval covered almost all the observations, which proves a measure of the reliability and robustness of the predictions. This study provides rich scientific references for applying advanced data-driven methods in time-series forecasting tasks and a practical methodological framework for water resources management and similar projects.

Enhancing The Performance of Intrusion Detection Using CNN And Reduction Techniques

There have been several security solutions based on artificial intelligence (AI), such as intrusion detection systems (IDS),cyberattacks are increasing because big data is increasing by using the internet on all sides of life, therefore, unbalanced data poses a serious problem in intrusion detection systems. The proposed detection system that is based on deep learning Convolutional Neural Network(CNN )partitions data into training and testing., Creating the classifier model for the Principal Component Analysis (PCA )technique of reducing features, is required for the development of intelligent analytic tools that need data pre-treatments and deep learning algorithm-performance enhancement. The UNSW-NB15 data set is used According to experimental findings, We employed a number of evaluation tools to assess the proposed NIDCNN strategy relying on the UNSW-NB15 data set that takes 30% of it for testing and after processing this part of the data became used to evaluate the proposed system. Measures such as a classifier's F-Score, precision, and sensitivity (Recall) are evaluated. Classifier performs better than the other approaches at determining if the data stream is normal or malicious. which is used to assess deep learning's effectiveness, the suggested model results from a high level of accuracy. The experimental findings demonstrate the suggested system's ability to accelerate the intrusion detection process while reducing memory and CPU usage. Experimental results prove the theoretical considerations.Because the UNSW-NB15 data set contains a wide range of patterns that accurately represent contemporary real network traffic, New NIDS algorithms can therefore be assessed using it.

Hyperparameters analysis of long short-term memory architecture for crop classification

<span lang="EN-US">Deep learning (DL) has seen a massive rise in popularity for remote sensing (RS) based applications over the past few years. However, the performance of DL algorithms is dependent on the optimization of various hyperparameters since the hyperparameters have a huge impact on the performance of deep neural networks. The impact of hyperparameters on the accuracy and reliability of DL models is a significant area for investigation. In this study, the grid Search algorithm is used for hyperparameters optimization of long short-term memory (LSTM) network for the RS-based classification. The hyperparameters considered for this study are, optimizer, activation function, batch size, and the number of LSTM layers. In this study, over 1,000 hyperparameter sets are evaluated and the result of all the sets are analyzed to see the effects of various combinations of hyperparameters as well the individual parameter effect on the performance of the LSTM model. The performance of the LSTM model is evaluated using the performance metric of minimum loss and average loss and it was found that classification can be highly affected by the choice of optimizer; however, other parameters such as the number of LSTM layers have less influence.</span>