Low False Detection Rate Research Articles

Video SAR Moving Target Detection Method Based on Machine Vision

When detecting moving targets in video SAR, it is impossible to accurately obtain the moving targets in dynamic images. In order to solve the problems of low detection integrity and poor effectiveness, a video SAR moving target detection method based on machine vision is proposed. Firstly, the moving target pixel points are matched by the mixed Gaussian model to obtain the target area of the video SAR moving image; secondly, the HSV model and reflectivity algorithm are used to remove the shadows caused by the moving target; finally, the processed target area is input into the YOLO algorithm to complete the final video SAR moving target detection. Experimental results show that the proposed algorithm has high detection integrity, high detection rate, low false detection rate, and better detection effectiveness.

A robust detection scheme against selective forwarding attacks in wireless sensor networks under variable harsh environments

Event-driven wireless sensor networks (EWSNs) often operate in harsh environments and are susceptible to selective forwarding attacks, where malicious nodes selectively drop packets. These attacks are difficult to detect due to similarities with normal node behavior under adverse conditions. To address this, we propose a detection scheme combining Variable Autoencoder (VAE) and Gaussian Mixture Model (GMM) with an Autoencoder (AE). The scheme extracts features from the node's single round forwarding rate (SFR), with VAE generating latent variables and GMM clustering them to differentiate behaviors. The AE model, trained on data representing normal behavior, uses reconstruction error to identify malicious nodes. Simulation results show that our scheme improves network throughput and achieves low False Detection Rate (FDR) and Missed Detection Rate (MDR), with both rates as low as 0.5% in normal conditions and below 2% in mobile harsh environments. With a computational complexity of O(n), this method is suitable for resource-constrained environments, demonstrating superior robustness compared to existing approaches.

COMPARISON OF MACHINE LEARNING ALGORITHMS FOR DETECTION OF DATA EXFILTRATION OVER DNS

Nowadays, computers are indispensable for business processes and home users. The widespread use of the Internet provides convenience in many areas from education to research. However, most of the users are unaware of technical security measures and use the Internet unconsciously. This situation leads to inadequate security measures against cyber-attacks. Various trainings are organised for conscious and safe internet use, but these efforts are not enough. Therefore, artificial intelligence-based solutions that can detect cyber incidents and close security gaps are becoming necessary. DNS tunnelling is a method used by malware to leak data over the internet. Vulnerable computers can put users in difficult situations by learning IP addresses from the wrong DNS servers. Innovative methods have been developed to detect this tunnelling. Some methods can detect low and slow data leakage through DNS in real time. There are also hybrid DNS tunnelling detection systems that achieve high accuracy and F-score using packet length and specific features. Feature-based methods sensitive to cache characteristics effectively characterise DNS tunnelling traffic with low false detection rates. These methods offer effective strategies for internet security. In this study, the detection of DNS tunnelling attacks by machine learning algorithms on the CIC-Bell-DNS-EXF-2021 dataset was investigated.

COTTON-YOLO: Enhancing Cotton Boll Detection and Counting in Complex Environmental Conditions Using an Advanced YOLO Model

This study aims to enhance the detection accuracy and efficiency of cotton bolls in complex natural environments. Addressing the limitations of traditional methods, we developed an automated detection system based on computer vision, designed to optimize performance under variable lighting and weather conditions. We introduced COTTON-YOLO, an improved model based on YOLOv8n, incorporating specific algorithmic optimizations and data augmentation techniques. Key innovations include the C2F-CBAM module to boost feature recognition capabilities, the Gold-YOLO neck structure for enhanced information flow and feature integration, and the WIoU loss function to improve bounding box precision. These advancements significantly enhance the model’s environmental adaptability and detection precision. Comparative experiments with the baseline YOLOv8 model demonstrated substantial performance improvements with COTTON-YOLO, particularly a 10.3% increase in the AP50 metric, validating its superiority in accuracy. Additionally, COTTON-YOLO showed efficient real-time processing capabilities and a low false detection rate in field tests. The model’s performance in static and dynamic counting scenarios was assessed, showing high accuracy in static cotton boll counting and effective tracking of cotton bolls in video sequences using the ByteTrack algorithm, maintaining low false detections and ID switch rates even in complex backgrounds.

Sample-efficient safety assurances using conformal prediction

When deploying machine learning models in high-stakes robotics applications, the ability to detect unsafe situations is crucial. Early warning systems can provide alerts when an unsafe situation is imminent (in the absence of corrective action). To reliably improve safety, these warning systems should have a provable false negative rate; that is, of the situations that are unsafe, fewer than ϵ will occur without an alert. In this work, we present a framework that combines a statistical inference technique known as conformal prediction with a simulator of robot/environment dynamics, in order to tune warning systems to provably achieve an ϵ false negative rate using as few as 1/ϵ data points. We apply our framework to a driver warning system and a robotic grasping application, and empirically demonstrate the guaranteed false negative rate while also observing a low false detection (positive) rate.

Trust evaluation model of power terminal based on equipment portrait

As the number of power terminals continues to increase and their usage becomes more widespread, the security of power systems is under great threat. In response to the lack of effective trust evaluation methods for terminals, we propose a trust evaluation model based on equipment portraits for power terminals. First, we propose an exception evaluation method based on the network flow order and evaluate anomalous terminals by monitoring the external characteristics of network traffic. Second, we propose an exception evaluation method based on syntax and semantics. The key fields of each message are extracted, and the frequency of keywords in the message is statistically analyzed to obtain the keyword frequency and time-slot threshold for evaluating the status of the terminal. Thus, by combining the network flow order, syntax, and semantic analysis, an equipment portrait can be constructed to guarantee security of the power network terminals. We then propose a trust evaluation method based on an equipment portrait to calculate the trust values in real time. Finally, the experimental results of terminal anomaly detection show that the proposed model has a higher detection rate and lower false detection rate, as well as a higher real-time performance, which is more suitable for power terminals.

A novel visual tracking scheme for unstructured indoor environments

In the ever-expanding sphere of assistive robotics, the pressing need for advanced methods capable of accurately tracking individuals within unstructured indoor settings has been magnified. This research endeavours to devise a realtime visual tracking mechanism that encapsulates high performance attributes while maintaining minimal computational requirements. Inspired by the neural processes of the human brain’s visual information handling, our innovative algorithm employs a pattern image, serving as an ephemeral memory, which facilitates the identification of motion within images. This tracking paradigm was subjected to rigorous testing on a Nao humanoid robot, demonstrating noteworthy outcomes in controlled laboratory conditions. The algorithm exhibited a remarkably low false detection rate, less than 4%, and target losses were recorded in merely 12% of instances, thus attesting to its successful operation. Moreover, the algorithm’s capacity to accurately estimate the direct distance to the target further substantiated its high efficacy. These compelling findings serve as a substantial contribution to assistive robotics. The proficient visual tracking methodology proposed herein holds the potential to markedly amplify the competencies of robots operating in dynamic, unstructured indoor settings, and set the foundation for a higher degree of complex interactive tasks.

Early and Consistent CRS Detection Using Wearable Device for Remote Patient Monitoring Following CAR-T Therapy in Relapsed/Refractory Multiple Myeloma (RRMM): Early Results of an Investigator-Initiated Trial

The administration of Chimeric Antigen Receptor T-cell (CAR-T) therapy for relapsed refractory multiple myeloma (RRMM) predominantly takes place in the inpatient setting to monitor for adverse effects like cytokine release syndrome (CRS) and neurotoxicity leading to increased healthcare utilization, infection risk and patient discomfort. Shifting the administration model to the outpatient setting may improve patient experience, optimize hospital bed utilization and decrease costs, but requires an objective and reliable method for CRS detection to monitor patients remotely and safely. The feasibility of using a wearable device for detecting CRS following CAR-T therapy in RRMM was compared to standard of care (SoC) nursing-recorded vital signs in this investigator initiated clinical trial (IIT). Patients wore a wearable device that collected continuous measures of temperature, pulse, respiratory rate, and O2 saturation in addition to SoC while receiving CAR-T therapy from the day infusion to discharge. Changes in vital signs captured by the device were analyzed based on 2 thresholds: (i) CRS timestamps were tagged in wearable data when pt's temperature breached a fixed threshold of 38 0C ( Tf) defined as fever by the American Society of Transplantation and Cellular Therapy (ASTCT), and (ii) when pt's temperature breached an individualized threshold of 2 standard deviations above their baseline temperature ( Ti) [see Figure 1]. Outcomes were time to first detection of CRS in wearable data vs. SoC. Wearable adherence was the duration pts wore the device over the total monitoring period. To date, 34 pts were screened, 28 enrolled (82.3% uptake), 1 was excluded from the dataset because of concurrent COVID-19, and 2 pts with CRS were excluded from time to detection comparison because of wearable adherence <50% during high-risk windows (defined as the time period where CRS is expected per known data for the CAR-T product). Max CRS grades were 1 (17 pts), 2 (1 pt), 3 (2 pts), and 4 (1 pt); 5 pts did not have clinical CRS. Patients wore the device for a median of 13 (12-15) days out of a median of 15.5 (14-17) days of inpatient admission for CAR-T infusion. Median individualized temperature threshold was 37.4 (37-37.6) C. Wearable adherence was 64 (51 - 77) % and 72 (57 - 87) % for the overall monitoring and high-risk periods, respectively. The wearable detected the initial CRS events at a median of 22 (-43 - 113.5) min earlier than SoC with Tf and a median of 184 (53 - 312) mins earlier than SoC with Ti method. In the no CRS subgroup, false CRS detection occurred in 1 pt with the Tf and 2 pts with the Ti method. Preliminary results suggest that time to CRS detection by the wearable device preceded SoC detection by a median of 103 minutes corroborated by both threshold methods. Temperature thresholds calibrated to the baseline temperature of individual patients facilitate earlier detection of CRS than fixed thresholds per clinical/ASTCT criteria. While the low false detection rate in is encouraging, these may also potentially be attributed to subclinical events. Both pts who had “false positive” triggers breached their individualized temperature thresholds via device readings while only recording low-grade temperatures that did not meet clinical CRS criteria per ASTCT, indicating possible capture of subclinical events. The study showcases the utility of continuous remote patient vital signs monitoring by wearable devices in efficiently capturing CRS, including subclinical events, earlier than SoC. Objective and reliable CRS monitoring may help safely transition workflows to outpatient CAR-T administration models. Analysis of trends in cytokine biomarkers for early CRS detection signal(s) is currently underway. Cytokine signals, alone and in combination with device data, are planned to be incorporated into machine learning models to build a CRS prediction index in future analyses. Figure 1: Snapshot in time of the continuous temperature data generated by the wearable device after CAR-T infusion showing capture of cytokine release syndrome. – Ti (solid line) represents the median baseline temperature of the patient over the entire duration of the data capture by the wearable device – Tf (dotted line) represents the temperature threshold of 38 0C, the clinical definition of fever as per ASTCT – Black diamond: clinical CRS event, when temperature >38 0C as recorded by nursing

RA-UNet: an intelligent fish phenotype segmentation method based on ResNet50 and atrous spatial pyramid pooling

Introduction: Changes in fish phenotypes during aquaculture must be monitored to improve the quality of fishery resources. Therefore, a method for segmenting and measuring phenotypes rapidly and accurately without harming the fish is essential. This study proposes an intelligent fish phenotype segmentation method based on the residual network, ResNet50, and atrous spatial pyramid pooling (ASPP).Methods: A sufficient number of fish phenotypic segmentation datasets rich in experimental research was constructed, and diverse semantic segmentation datasets were developed. ResNet50 was then built as the backbone feature extraction network to prevent the loss of fish phenotypic feature information and improve the precision of fish phenotypic segmentation. Finally, an ASPP module was designed to improve the phenotypic segmentation accuracy of different parts of fish.Results: The test algorithm based on the collected fish phenotype segmentation datasets showed that the proposed algorithm (RA-UNet) yielded the best results among several advanced semantic segmentation models. The mean intersection over union (mIoU) and mean pixel accuracy (mPA) were 87.8% and 92.3%, respectively.Discussion: Compared with the benchmark UNet algorithm, RA-UNet demonstrated improvements in the mIoU and mPA by 5.0 and 1.8 percentage points, respectively. Additionally, RA-UNet exhibited superior fish phenotype segmentation performance, with a low false detection rate and clear and complete edge segmentation. Conclusively, the RA-UNet proposed in this study has high accuracy and edge segmentation ability and can, therefore, directly improve the efficiency of phenotypic monitoring in fish farming.

Multi-Sensor Fusion for Proactive Anomaly Detection in Robot Navigation

Mobile robots frequently exhibit abnormal behaviors that can impair navigation despite the rapid progress of navigation algorithms. Modern robots need to be able to recognize these unusual behaviors in order to reach high levels of autonomy. Methods for reactive anomaly detection Detect anomalies poor task executions based on the current state of the robot and thus lack the capability to warn the robot before a malfunction actually happens. Due to the possibility of harm to the robot and the environment, such a warning delay is undesirable. For robot navigation in unstructured and uncertain situations, we suggest a proactive anomaly detection network (PAAD). Based on the anticipated movements from the predictive controller and the current observations from the perception module, PAAD forecasts the likelihood of future failure. Effective fusion of multi-sensor signals to provide reliable anomaly detection as seen in the field when there is sensor occlusion environments. Our tests on data from field robots show that our model can catch abnormal actions in real-time while retaining a low false detection rate in congested areas, outperforming earlier methods in failure identification IndexTerms - Face recognition, bias, fairness, soft-biometrics, analysis, privacy, biometrics

CXRNet: CNN‐attention based CXR image classifier

AbstractChest X‐ray (CXR) images are widely accepted for the diagnosis of lung diseases. The X‐ray machinery is widely available but the number of radiologists who interpret these images is very limited. Therefore, the development of an automated disease classification system is a need for the healthcare industry. The existing testing methods take hours to days to generate the testing result and have low detection accuracy and high false detection rate. Furthermore, the testing kits are costly, and availability is limited. Therefore, Convolutional Neural Network (CNN) based framework is proposed to address these limitations. Four pre‐trained frameworks, ResNet50V2, InceptionV3, NASNetMobile, and Xception are used to generate highly dense features, refine the features using the attention module, and then fused these features to classify the diseases using CXR images. In addition, extensive experiments were carried out on the activation functions (Relu, Leaky Relu, and Tanh), which help to improve the results. Evaluation is done on the three enriched CXR datasets DS‐1, DS‐2, and DS‐3 to examine the performance of the proposed framework in terms of binary classification and multi‐class classification. The proposed model achieved a class‐5 accuracy of 92.89% on dataset DS‐3 and the class‐3 classification accuracy of 95.22%, 92.25% on dataset DS‐1 and DS‐2, respectively. The class‐2 classification accuracy on dataset DS‐1 and DS‐2 is 99.23% and 98.85%, respectively.

LSTM-NV: A combined scheme against selective forwarding attack in event-driven wireless sensor networks under harsh environments

Wireless Sensor Networks (WSNs) are susceptible to selective forwarding attacks, which can lead to reduced network efficiency and compromise the integrity of transmitted data. The indistinguishability between malicious behavior and normal packet loss in harsh environments exacerbates the challenge. To address this, in this paper, we provide a combined detection scheme called LSTM-NV, consisting of a training stage and a detecting stage. During the training stage, we integrate variational mode decomposition (VMD) with a long short term memory (LSTM) model to learn normal nodes’ forwarding behavior and then predict errors for each node. During the detecting stage, dynamic thresholds are determined to identify local anomaly points and a novel neighbor voting method is employed to differentiate between malicious and normal nodes. Our scheme demonstrates superior performance with a low average missed detection rate (MDR) of 0.6% and a low average false detection rate (FDR) of 3.3% compared to other effective methods, while also offering lower detection algorithmic complexity.

Computerized Block Games for Automated Cognitive Assessment: Development and Evaluation Study

Cognitive assessment using tangible objects can measure fine motor and hand-eye coordination skills along with other cognitive domains. Administering such tests is often expensive, labor-intensive, and error prone owing to manual recording and potential subjectivity. Automating the administration and scoring processes can address these difficulties while reducing time and cost. e-Cube is a new vision-based, computerized cognitive assessment tool that integrates computational measures of play complexity and item generators to enable automated and adaptive testing. The e-Cube games use a set of cubes, and the system tracks the movements and locations of these cubes as manipulated by the player. The primary objectives of the study were to validate the play complexity measures that form the basis of developing the adaptive assessment system and evaluate the preliminary utility and usability of the e-Cube system as an automated cognitive assessment tool. This study used 6 e-Cube games, namely, Assembly, Shape-Matching, Sequence-Memory, Spatial-Memory, Path-Tracking, and Maze, each targeting different cognitive domains. In total, 2 versions of the games, the fixed version with predetermined sets of items and the adaptive version using the autonomous item generators, were prepared for comparative evaluation. Enrolled participants (N=80; aged 18-60 years) were divided into 2 groups: 48% (38/80) of the participants in the fixed group and 52% (42/80) in the adaptive group. Each was administered the 6 e-Cube games; 3 subtests of the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV; Block Design, Digit Span, and Matrix Reasoning); and the System Usability Scale (SUS). Statistical analyses at the 95% significance level were applied. The play complexity values were correlated with the performance indicators (ie, correctness and completion time). The adaptive e-Cube games were correlated with the WAIS-IV subtests (r=0.49, 95% CI 0.21-0.70; P<.001 for Assembly and Block Design; r=0.34, 95% CI 0.03-0.59; P=.03 for Shape-Matching and Matrix Reasoning; r=0.51, 95% CI 0.24-0.72; P<.001 for Spatial-Memory and Digit Span; r=0.45, 95% CI 0.16-0.67; P=.003 for Path-Tracking and Block Design; and r=0.45, 95% CI 0.16-0.67; P=.003 for Path-Tracking and Matrix Reasoning). The fixed version showed weaker correlations with the WAIS-IV subtests. The e-Cube system showed a low false detection rate (6/5990, 0.1%) and was determined to be usable, with an average SUS score of 86.01 (SD 8.75). The correlations between the play complexity values and performance indicators supported the validity of the play complexity measures. Correlations between the adaptive e-Cube games and the WAIS-IV subtests demonstrated the potential utility of the e-Cube games for cognitive assessment, but a further validation study is needed to confirm this. The low false detection rate and high SUS scores indicated that e-Cube is technically reliable and usable.

A Small Sample Photovoltaic Hot Spot Identification Method Based on Deep Transfer Learning

A small-sample photovoltaic hot spot identification method based on deep transfer learning has been proposed as a solution to the problem that traditional deep learning models require a substantial amount of training data, whereas the number of hot spot effect samples is relatively low and difficult to collect. This issue can be resolved by the development of a solution that employs deep transfer learning. The Inception-v 4 model serves as the basis for building a deep transfer learning model. After finishing the training procedure using a small-sample hot spot dataset of negative sample multi-classification, a network model for hot spot detection is built. Experiments reveal that even when the number of accessible samples is restricted, the model trained using the deep transfer learning technique has great identification accuracy, a low false detection rate, and strong generalization capability.

Intrusion Detection-Data Security Protection Scheme Based on Particle Swarm-BP Network Algorithm in Cloud Computing Environment

Aiming at the problems of low detection rate and high false detection rate of intrusion detection algorithms in the traditional cloud computing environment, an intrusion detection-data security protection scheme based on particle swarm-BP network algorithm in a cloud computing environment is proposed. First, based on the four modules of data collection, data preprocessing, feature selection, and intrusion detection, the overall framework of the intrusion detection model is constructed by designing corresponding functions. Then, by introducing the decision tree algorithm, the overfitting is reduced and the data processing speed of the model is improved, and on this basis, the feature selection is carried out through the “gain rate” optimization method, which reduces the redundant information of the feature vector. Finally, by introducing the Particle Swarm Optimization (PSO) algorithm into the optimization of the initial weights and thresholds of the BP neural network, the BP neural network is improved based on the momentum factor and adaptive learning rate, and the high detection rate and low false detection rate are realized. Through simulation experiments, the proposed intrusion detection method and the other three methods are compared and analyzed under the same conditions. The results show that the detection rate and false detection rate of the method proposed in this paper are the best under five different types of sample data, the highest detection rate reaches 95.72%, and the lowest false detection rate drops to 2.03%. The performance of the proposed algorithm is better than that of the other two comparison algorithms.

A Network Intrusion Detection Method Based on Improved Bi-LSTM in Internet of Things Environment

When performing malicious network attack detection, traditional intrusion detection methods show their disadvantage of low accuracy and high false detection rate. To address these problems, this paper proposes a novel network intrusion detection scheme based on an improved bi-directional long short-term memory (Bi-LSTM) model under the emerging internet of things (IoT) environment. Firstly, this paper analyzes Bi-LSTM model. Then, it introduces a two-layer attention network structure into Bi-LSTM network. Finally, the corresponding network intrusion detection system is constructed based on the improved Bi STM model. Through simulation experiments, the proposed network intrusion detection method and other three methods are compared under five identical databases. Experimental results show that the false detection rate and detection accuracy of the proposed method are optimal on all sample data, the detection accuracy reaches 97.24% and the false detection rate drops to 5.13%.

Uncertainty Quantification for Full-Flight Data Based Engine Fault Detection with Neural Networks

Current state-of-the-art engine condition monitoring is based on a minimum of one steady-state data point per flight. Due to the scarcity of available data points, there are difficulties distinguishing between random scatter and an underlying fault introducing a detection latency of several flights. Today’s increased availability of data acquisition hardware in modern aircraft provides continuously sampled in-flight measurements, so-called full-flight data. These full-flight data give access to sufficient data points to detect faults within a single flight, significantly improving the availability and safety of aircraft. Artificial neural networks are considered well suited for the timely analysis of an extensive amount of incoming data. This article proposes uncertainty quantification for artificial neural networks, leading to more reliable and robust fault detection. An existing approach for approximating the aleatoric uncertainty was extended by an Out-of-Distribution Detection in order to take the epistemic uncertainty into account. The method was statistically evaluated, and a grid search was performed to evaluate optimal parameter combinations maximizing the true positive detection rates. All test cases were derived based on in-flight measurements of a commercially operated regional jet. Especially when requiring low false positive detection rates, the true positive detections could be improved 2.8 times while improving response times by approximately 6.9 compared to methods only accounting for the aleatoric uncertainty.

Multi-sensor target data detection method based on improved deep learning

Aiming at the problems of low detection accuracy and long detection time in the traditional multi-sensor target data detection method, a multi-sensor target data detection method based on improved deep learning is proposed. A regional nomination network model is built to locate all rectangular areas where multi-sensor targets may exist. A sample acquisition model is established by fusing multi-sensor target data in this area in combination with the storage structure of multi-sensor target data to collect multi-sensor target data and filter the collected target data. The histogram feature method is used to extract the processed multi-sensor target data features and the extracted multi-sensor target data features are input into the neural network training samples. The square error function is used to calculate the sample deviation and the deviation is corrected by the deviation gradient sinking method. Finally, the multi-sensor target data detection results are output. Simulation results show that the proposed method has high accuracy, low false detection rate and short detection time along with applicability.

The Detection Scheme Against Selective Forwarding of Smart Malicious Nodes With Reinforcement Learning in Wireless Sensor Networks

Wireless sensor networks (WSNs) are extremely vulnerable to different attacks because of open communication, and distribution in unattended areas. The selective forwarding attack is one of the most difficult inside attacks to be detected for two reasons. The node in a harsh environment has to drop some data packets, and the smart malicious node frequently eludes detection. In this paper, we model a selective forwarding attack of smart malicious nodes with a reinforcement learning (RL) algorithm. To effectively detect the selective forwarding attack under a harsh environment, we design the double-threshold density peaks clustering (DT-DPC) algorithm. Abnormal nodes are identified as malicious and isolated owing to continuous abnormalities. Suspicious nodes are determined by the neighbor voting method because malicious behaviors show up separately and a harsh environment universally disturbs agglomerate nodes. Even if smart malicious nodes elude the detection by an RL algorithm, DT-DPC improves the network throughput. The simulation results show that DT-DPC has a low false detection rate (FDR) of around 1% and a missed detection rate (MDR) of around 10%. The network throughput increases by about 4% under a harsh environment.

Intelligent Cyber Security Framework Based on SC-AJSO Feature Selection and HT-RLSTM Attack Detection

Cyber security is identified as an emerging concern for information technology management in business and society, owing to swift advances in telecommunication and wireless technologies. Cyberspace security has had a tremendous impact on numerous crucial infrastructures. Along with current security status data, historical data should be acquired by the system to implement the latest cyber security defense and protection. It also makes intelligent decisions that can provide adaptive security management and control. An intelligent cyber security framework using Hyperparameter Tuning based on Regularized Long Short-Term Memory (HT-RLSTM) technique was developed in this work to elevate the security level of core system assets. To detect various attacks, the proposed framework was trained and tested on the collection of data. Owing to missing values, poor scaling, imbalanced and overlapped data, the data was primarily incomplete and inconsistent. To elevate the decision making for detecting attacks, the inconsistent or unstructured data issue was addressed. The missing values were handled by this work along with scaling performance using the developed Kernelized Robust Scaler (KRS). Using the developed Random Over Sample-Based Density-Based Spatial Clustering Associated with Noise (ROS-DBSCAN), the imbalanced and overlapped data were handled, which was followed by the relevant feature selection of data utilizing the Sine Cosine-Based Artificial Jellyfish Search Optimization (SC-AJSO) technique. The data were split under the provision of Stratified K-Fold cross-validation along being trained in the proposed HT-RLSTM. The experimental analysis depicted that better accuracy was attained in detecting attacks by the proposed work for different datasets. When analogized with prevailing state-of-the-art methods, a low false detection rate, as well as computation time, was attained by the proposed scheme.