Occurrence Of False Positives Research Articles

A comparative analysis of convolutional neural networks for breast cancer prediction

Breast cancer continues to be a substantial worldwide health concern, affecting millions of individuals each year; this emphasizes the critical nature of early detection in order to enhance patient prognoses. The present study aims to assess the classification performance of three convolutional neural network (CNN) architectures-visual geometry group 19 (VGG19), AlexNet, and residual network 50 (ResNet50)-with respect to breast cancer detection in medical images. Thorough assessments, encompassing metrics such as accuracy, precision, recall, and F-score, were undertaken to evaluate the diagnostic performance of the models. ResNet50 consistently outperforms other models, as evidenced by its highest accuracy and F-score. The research highlights the significant importance of carefully choosing suitable architectures for medical image analysis, with a specific focus on the detection of breast cancer. In addition, it demonstrates the capacity of deep learning models, such as ResNet50, to improve the diagnosis of breast cancer with exceptional precision and sensitivity, which is critical for reducing the occurrence of false positives and negatives in clinical environments. In addition, computational efficiency is taken into account; AlexNet is recognized as the most efficient model, which is advantageous in environments with limited resources. This study advances medical image processing by demonstrating the potential of CNNs in the detection of breast cancer. The results of this study establish a fundamental basis for sub- sequent inquiries and suggest approaches to improve timely detection and treatment, which will ultimately be advantageous for both patients and healthcare professionals.

A falls recognition framework based on faster R-CNN and temporal action sequences that can be deployed on home service robots

With the popularity of home service robots (e.g. floor sweepers), robots should be considered to have more features for older adult care. Compared to fixed home monitors with a limited field of view, fall detection with service robots is an ideal solution to keep older adults and disabled people within sight. However, the user’s actions, such as lying on the bed to sleep or slumping on the sofa to rest, cause the traditional fall detection system to generate false alarms, which disrupts the user’s family life. The present work proposed an enhanced faster R-convolutional neural network (CNN) network by incorporating temporal action sequences and fall acceleration computation, demonstrating a reduced misjudgment rate on the service robot platform. Firstly, motion images were captured to obtain the target’s motion area description and action timing at the input stage. Then, the faster R-CNN algorithm was implemented to check the suspected falls further based on the falling acceleration of the detected actions during the training phase. Finally, the proposed temporal action sequences module eliminated the action mistaken for falling. Network training and robotic platform testing demonstrated that the proposed approach distinguished between falls and false alarms, which mitigated the occurrence of false positives. On the service robot platform, experimental results showed that the FAR was 8.19 and processing time was 0.79 s.

Improved object detection method for unmanned driving based on Transformers.

The object detection method serves as the core technology within the unmanned driving perception module, extensively employed for detecting vehicles, pedestrians, traffic signs, and various objects. However, existing object detection methods still encounter three challenges in intricate unmanned driving scenarios: unsatisfactory performance in multi-scale object detection, inadequate accuracy in detecting small objects, and occurrences of false positives and missed detections in densely occluded environments. Therefore, this study proposes an improved object detection method for unmanned driving, leveraging Transformer architecture to address these challenges. First, a multi-scale Transformer feature extraction method integrated with channel attention is used to enhance the network's capability in extracting features across different scales. Second, a training method incorporating Query Denoising with Gaussian decay was employed to enhance the network's proficiency in learning representations of small objects. Third, a hybrid matching method combining Optimal Transport and Hungarian algorithms was used to facilitate the matching process between predicted and actual values, thereby enriching the network with more informative positive sample features. Experimental evaluations conducted on datasets including KITTI demonstrate that the proposed method achieves 3% higher mean Average Precision (mAP) than that of the existing methodologies.

No tricks no bluff, focusing on localizing crisp boundaries in image media

Boundary detection, as a fundamental task for computer vision applications, plays an important role in many tasks such as image deblurring, semantic segmentation, camouflaged object detection, and salient object detection. The thickness problem of boundary prediction is so prevalent in existing methods that NMS has to be used as a post-processing method to thin the prediction boundaries. However, we believe that additional post-processing does not constitute an end-to-end behavior, and the results obtained do not reflect the true boundary localization capabilities of the algorithm. In comparison to the superior end-to-end approach, NMS reveals notable limitations in rectifying erroneous predictions and achieving high-quality results similar to the ground truth. Furthermore, the use of NMS will inevitably compromise the prediction of fragile boundaries. Therefore, we remove the NMS step from the boundary detection process and generalize two reasons for the detector’s poor ability to localize crisp boundaries: boundary blurring and static cues. Aliasing region loss (Lar) and hollow region loss (Lhr) are introduced to mitigate the occurrence of false positives in various types of regions, thereby addressing the issue of boundary blurring. For static cues, we propose a dimension attention (DimAttn) module which can realize multidimensional and multi-range clew interaction. In fact, our method can be plug-and-play in any existing boundary detection model. Extensive experiments have demonstrated that the boundary prediction results of the proposed framework are crisper and exhibit remarkable superiority in detecting fragile boundaries compared with scenarios utilizing NMS.

Optimizing Spectronaut Search Parameters to Improve Data Quality with Minimal Proteome Coverage Reductions in DIA Analyses of Heterogeneous Samples.

Data-independent acquisition has seen breakthroughs that enable comprehensive proteome profiling using short gradients. As the proteome coverage continues to increase, the quality of the data generated becomes much more relevant. Using Spectronaut, we show that the default search parameters can be easily optimized to minimize the occurrence of false positives across different samples. Using an immunological infection model system to demonstrate the impact of adjusting search settings, we analyzed Mus musculus macrophages and compared their proteome to macrophages spiked withCandida albicans. This experimental system enabled the identification of "false positives" as Candida albicans peptides and proteins should not be present in the Mus musculus-only samples. We show that adjusting the search parameters reduced "false positive" identifications by 89% at the peptide and protein level, thereby considerably increasing the quality of the data. We also show that these optimized parameters incurred a moderate cost, only reducing the overall number of "true positive" identifications across each biological replicate by <6.7% at both the peptide and protein level. We believe the value of our updated search parameters extends beyond a two-organism analysis and would be of great value to any DIA experiment analyzing heterogeneous populations of cell types or tissues.

Abstract 5018: Evaluation of whole blood fractionation conditions for optimal liquid biopsy NGS applications

Abstract “For research use only. Not for diagnostic procedures.” Introduction: Liquid biopsy is a non-invasive alternative to tissue biopsies for cancer profiling using cell-free total nucleic acids (cfTNA; including cell free DNA (cfDNA) and cell free RNA (cfRNA)), the biological analytes derived from tumor cells floating in blood. However, it is crucial to optimize the whole blood fractionation workflows, since limited access to a suitable centrifuge at blood collection sites poses a challenge to this critical fractionation step, causing delays in processing the blood, leading to decreased sensitivity of variant detection and increased occurrence of false positives. Current study shows a systematic comparison of the effects of 8 different blood fractionation conditions for spin speeds, temperatures and duration on the levels of cfTNA derived from K2-EDTA-derived blood from multiple healthy and non-small cell lung cancer (NSCLC) donors, using next generation sequencing (NGS). Methods: CfTNA from K2-EDTA derived plasma was isolated using the MagMAXTM Cell-Free Total Nucleic Acid kit and the Genexus™ Cell free total nucleic acid kit, utilizing the KingFisher TM Flex and GenexusTMpurification systems. Yield and integrity of cfTNA isolated was evaluated by Qubit™, Agilent Tapestation and quantitative real time PCR (qRT-PCR). Furthermore, the impact of these test conditions on quality control specifications of DNA and RNA was assessed by performing sequencing on an Ion TorrentTM GenexusTM sequencing instrument using the AmpliSeqTM HD target amplification assay, that reports sequence variations in DNA for 42 cancer related genes and RNA in 7 cancer related genes. Results: Results showed comparable performance for all the 8 blood fractionation conditions tested, for yield and cfDNA profiles compared to control condition, indicating that there is no significant impact of these conditions on either yield or integrity of the cfDNA isolated. In addition, NGS data also showed that on target reads and median molecular coverage for DNA are comparable for all the conditions tested. However, sequencing data demonstrated that one test condition showed significantly better performance for RNA quality control specifications in terms of RNA mapped reads and molecular coverage for process controls used in the assay, compared to all other conditions tested. Conclusion: Data suggests that one of the whole blood fractionation conditions tested impact the cfRNA recovery from plasma without negatively affecting the yield, and integrity of isolated cfDNA and shows increased sensitivity for RNA process controls in NGS. Citation Format: MADHU JASTI, Sergio Hernandez, Nick Siepert, Lokesh Kodandaramaswamy, Thilanka Jayaweera, Jeoffrey Schageman, Angie Cheng, Susanne Lopez, Miao Der Chen. Evaluation of whole blood fractionation conditions for optimal liquid biopsy NGS applications [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5018.

Distribution of Signal to Noise Ratio and Application to Leakage Detection

In the context of side-channel attacks, the Signal to Noise Ratio (SNR) is a widely used metric for characterizing the information leaked by a device when handling sensitive variables. In this paper, we derive the probability density function (p.d.f.) of the signal to noise ratio (SNR) for the byte value and Hamming Weight (HW) models, when the number of traces per class is large and the target SNR is small. These findings are subsequently employed to establish an SNR threshold, guaranteeing minimal occurrences of false positives. Then, these results are used to derive the theoretical number of traces that are required to remain below pre-defined false negative and false positive rates. The sampling complexity of the T-test, ρ-test and SNR is evaluated for the byte value and HW leakage model by simulations and compared to the theoretical predictions. This allows to establish the most pertinent strategy to make use of each of these detection techniques.

ESSENCE-Dock: A Consensus-Based Approach to Enhance Virtual Screening Enrichment in Drug Discovery.

Drug development is a complex, costly, and time-consuming endeavor. While high-throughput screening (HTS) plays a critical role in the discovery stage, it is one of many factors contributing to these challenges. In certain contexts, virtual screening can complement the HTS, potentially offering a more streamlined approach in the initial stages of drug discovery. Molecular docking is an example of a popular virtual screening technique that is often used for this purpose; however, its effectiveness can vary greatly. This has led to the use of consensus docking approaches that combine results from different docking methods to improve the identification of active compounds and reduce the occurrence of false positives. However, many of these methods do not fully leverage the latest advancements in molecular docking. In response, we present ESSENCE-Dock (Effective Structural Screening ENrichment ConsEnsus Dock), a new consensus docking workflow aimed at decreasing false positives and increasing the discovery of active compounds. By utilizing a combination of novel docking algorithms, we improve the selection process for potential active compounds. ESSENCE-Dock has been made to be user-friendly, requiring only a few simple commands to perform a complete screening while also being designed for use in high-performance computing (HPC) environments.

Empowering Independence through Real Time Object Identification and Navigation for People with Disabilities

Recent studies in assistive technologies for visually impaired individuals showcase a diverse range of methodologies, algorithms, and implementations aimed at enhancing their independence. A notable focus revolves around leveraging cutting-edge technologies such as YOLO (You Only Look Once), SSD (Single Shot Multibox Detector), and Faster R-CNN (Region-based Convolutional Neural Network) to develop real-time object detection systems and deep learning-based smartphone navigation solutions . One prevalent theme in these advancements is the incorporation of auditory feedback to facilitate enhanced user interaction. This is achieved through sophisticated text-to-speech conversion and the integration of audio cues. The utilization of auditory cues not only aids in real-time awareness of the surroundings but also significantly contributes to the overall user experience . Despite remarkable progress, challenges persist in the realm of assistive technologies for the visually impaired. Issues such as processing speed, the occurrence of false positives and negatives, and the adaptability of these systems to various environmental conditions remain prominent. These challenges underline the need for continued research and development in this field to address existing limitations and refine the effectiveness of these assistive technologies .In essence, this survey provides a comprehensive understanding of the current landscape of assistive technologies for the visually impaired. By identifying both achievements and existing challenges, it serves as a valuable resource for researchers and practitioners, contributing to ongoing advancements that ensure tailored solutions and improved independence for individuals with visual impairments

Dual-Signal Chemical Exchange Saturation Transfer (Dusi-CEST): An Efficient Strategy for Visualizing Drug Delivery Monitoring in Living Cells.

We report a dual-signal chemical exchange saturation transfer (Dusi-CEST) strategy for drug delivery and detection in living cells. The two signals can be detected by operators in complex environments. This strategy is demonstrated on a cucurbit[6]uril (CB[6]) nanoparticle probe, as an example. The CB[6] probe is equipped with two kinds of hydrophobic cavities: one is found inside CB[6] itself, whereas the other exists inside the nanoparticle. When the probe is dispersed in aqueous solution as part of a hyperpolarized 129Xe NMR experiment, two signals appear at two different chemical shifts (100 and 200 ppm). These two resonances correspond to the NMR signals of 129Xe in the two different cavities. Upon loading with hydrophobic drugs, such as paclitaxel, for intracellular drug delivery, the two resonances undergo significant changes upon drug loading and cargo release, giving rise to a metric enabling the assessment of drug delivery success. The simultaneous change of Dusi-CEST likes a mobile phone that can receive both LTE and Wi-Fi signals, which can help reduce the occurrence of false positives and false negatives in complex biological environments and help improve the accuracy and sensitivity of single-shot detection.

Machine Learning for Anomaly Detection in Electric Transportation Networks

This study introduces a sophisticated anomaly detection system based on machine learning. The system is specifically developed to enhance the dependability and safeguard the security of electric transportation networks, with a particular emphasis on the charging infrastructure for electric vehicles (EVs). Utilizing extensive datasets, the research examines several facets of charging stations, charging records, identified abnormalities, and following maintenance measures. The examination of the charging station demonstrates the system’s versatility in accommodating many charging circumstances, as seen by the range of power ratings, consumption patterns, and energy provided. Further examination of charging records provides comprehensive understanding of individual charging sessions, enabling the detection of irregularities such as atypical energy surges and extended charging durations. The machine learning system, having been trained and verified using this data, has a commendable degree of precision in identifying anomalies, as shown by the congruence between anticipated abnormalities and real results. The maintenance and repair measures carried out in reaction to identified abnormalities highlight the practical ramifications of the system, with proactive tactics utilized to reduce downtime and enhance charging station operations. The performance measures, including accuracy, recall, and F1 score, unequivocally validate the resilience of the anomaly detection system, guaranteeing precise identification while mitigating the occurrence of false positives and negatives. The seamless incorporation of machine learning into electric transportation networks, as shown by the results, not only amplifies the dependability and safeguarding of EV charging infrastructure but also establishes the system as an invaluable instrument for practical implementations. The research, in addition to offering a thorough examination of the system’s performance, elucidates forthcoming avenues for scalability, real-time monitoring, and interpretability, thereby making a valuable contribution to the wider discussion on the revolutionary capabilities of machine learning in the ever-changing realm of electric transportation.

The Genetic Diversity of Horse Native Breeds in Russia.

Horses were domesticated later than other farm animals. Horse breeds have been selectively developed by humans to satisfy different needs and purposes. The factory and indigenous breeds are of particular interest, having been bred in purity for many centuries without the addition of foreign blood. Data from 31 stud farms, as well as ranches, located in fifteen regions of the Russian Federation were used in this work. DNA was sampled from 102 stallions of 11 breeds: Arabian, Akhal-Teke, Don, Orlov Trotter, Vladimir Heavy Draft, Russian Heavy Draft, Soviet Heavy Draft, Kabardin, Yakut, Tuva, and Vyatka. Data on the origin of each animal from which the material was collected were taken into account. DNA genotyping was carried out using GGP Equine 70 k ® array chips (Thermo Fisher Scientific, USA). Genetic diversity of horse breeds was estimated using Admixture 1.3. and PLINK 1.9 software. FROH inbreeding was computed via the R detectRUNS package. The minimum length for ROH was set at 1 Mb to reduce the occurrence of false positives. We conducted PCA analysis using PLINK 1.9, and used the ggplot2 library in R for visualizing the results. Indigenous equine breeds, such as Vyatka, Tuva, and Yakut, are very hardy, and well adapted to local environmental and climatic conditions. They are employed as draft power, as well as for milk and meat. Both the Akhal-Teke breed and the Arabian breed have retained a minimum effective population size over many generations. We note significant accumulations of homozygosity in these breeds. In equestrian sports, performance is a top priority. ADMIXTURE and PCA analyses showed similarities between Don equine breeds and Kabardin, as well as some Arabian breed animals. Earlier research indicated the presence of thoroughbred traits in Don stallions. The Orlov Trotter breed stands out as a separate cluster in the structural and PCA analyses. Considering the small population size of this breed, our study found high FROH in all tested animals. The general reduction in the diversity of the horse breed gene pool, due to numerous crosses for breed improvement with thoroughbreds, has lead to a decline in the differences between the top sporting breeds. Our study presents new opportunities for exploring the genetic factors that influence the formation of adaptive traits in indigenous breeds, and for finding ways to preserve genetic diversity for effective population reproduction.

Machine-Learning-Based Spam Mail Detector

The proliferation of spam emails, a predominant form of online harassment, has elevated the significance of email in daily life. As a consequence, a substantial portion of individuals remain vulnerable to fraudulent activities. Despite Gmail’s “spam mail filtration system,” its effectiveness is not absolute. It occasionally misclassifies legitimate messages, leading to their confinement in the spam folder, or overlooks potentially harmful spam emails. This results in the occurrence of false positives. This research scrutinizes the historical data, cookies, caches, Session Restores, flash artifacts, and super cookies of Internet Explorer, Firefox, and Chrome on the Windows 10 platform. Data was collected through Google, Firefox, and Internet Explorer, operating within the Windows 10 environment. It has been observed that browsers store user behavior data on the host computer’s hard drive. The implications of this study hold substantial value for computer forensics researchers, law enforcement professionals, and digital forensics experts. The study leverages Python, alongside pertinent libraries such as pandas, Numpy, Matplotlib, scikit-learn, and flask, to facilitate its investigation. The experiment result and analysis show KN and NB algorithms have the best accuracy and precision score compared to other Algorithms.

Determination of seven mycotoxins in cereals by ultra performance liquid chromatography-quadrupole-time of flight mass spectrometry based on the self-built database

A method for the determination of seven mycotoxins in rice and wheat by ultra performance liquid chromatography-quadrupole-time of flight mass spectrometry (UPLC-Q-TOF/MS) based on self-built database was established. Samples were extracted with 0.2% formic acid aqueous solution-acetonitrile (50∶50, v/v), dehydrated and salted according to the QuEChERS method (4 g of magnesium sulfate, 1 g of sodium chloride, 1 g of sodium citrate, 0.5 g of citrate disodium salt), and separated on an HSS T3 column (100 mm×2.1 mm, 1.8 μm). UPLC-Q-TOF/MS with MSE screening was performed, and the positive and negative ions of the screened mycotoxins were calibrated and quantified using matrix-matched standard curves with time of flight multiple reaction monitoring (TOF-MRM). The results showed that aflatoxin G1 (AFG1), aflatoxin G2 (AFG2), aflatoxin B1 (AFB1), aflatoxin B2 (AFB2), and ochratoxin A (OTA) exhibited moderate matrix effects in rice, while OTA and zearalenone (ZEN) exhibited moderate matrix effects in wheat. The seven mycotoxins showed good linearities in their respective concentration ranges, with correlation coefficients (r) of 0.9900-0.9998. The limits of detection (LODs) for rice and wheat were 0.50-400 and 0.50-200 μg/kg, respectively, and the limits of quantification (LOQs) for both cereals were 1.00-800 μg/kg. In rice, the average recoveries at three spiked levels of low, medium, and high were 88.1%-123.9%, with relative standard deviations (RSDs) of 0.2%-13.6%. In wheat, the average recoveries at three spiked levels of low, medium, and high were 102.0%-123.4%, with RSDs of 0.8%-14.8%. As a result, one batch of 46 batches of rice was screened out for AFB1 and AFB2, with a screening rate of 2.2%, of which the measured values were 10.8 μg/kg and 1.2 μg/kg, respectively. According to GB 2761-2017, the maximum allowable level of AFB1 in rice is 10 μg/kg; thus, the exceeding rate for AFB1 is 2.2%. Deoxynivalenol (DON) was screened out in 9 out of 24 batches of wheat (screening rate, 37.5%), while ZEN was screened out in 19 batches (screening rate, 79.2%). According to GB 2761-2017, the maximum allowable levels of DON and ZEN in wheat are 1000 and 60 μg/kg, respectively. The levels of DON and ZEN detected in the wheat samples did not exceed these limits. The proposed method uses MSE for qualitative screening to avoid the occurrence of false positives caused by interfering compounds with mass numbers and retention times similar to those of the analytes. TOF-MRM mode is then used to quantify the positively screened mycotoxins. The method is fast, accurate, sensitive, and suitable for the isolation and quantitative detection of mycotoxin residues in rice and wheat samples. The findings provide powerful technical support for mycotoxin contamination monitoring in rice and wheat and early risk-warning efforts.

Test performance study to evaluate diagnostic tests for the detection and identification of Xylophilus ampelinus in vine material spiked with cells from X. ampelinus cultures

A test performance study (TPS) was conducted in 2020 to evaluate performance of serological and PCR-based tests for the detection of Xylophilus ampelinus in homogenised vine samples. In total, 11 labs participated, although there were fewer participants for the serological tests than the PCR-based tests. The panel of samples sent to participants included spiked samples containing 104–108 cfu per ml (serological tests) or 102–106 cfu per ml (PCR-based tests) as well as positive and negative controls. The five PCR-based tests were found to be fit for purpose, with similar performance across a range of metrics (analytical sensitivity, diagnostic sensitivity and specificity, and repeatability and reproducibility assessed in terms of accordance and concordance, respectively). Serological methods (two immunofluorescence tests and two ELISA tests) were found to be less sensitive with regard to both analytical and diagnostic sensitivity. Furthermore, the occurrence of false positives suggests that a positive IF result may not be conclusive when considered in isolation. One of the ELISA tests exhibited much lower analytical and diagnostic sensitivity than the other serological tests and would not be considered suitable for the purpose considered by this TPS.

Cancer care at the time of the fourth industrial revolution: an insight to healthcare professionals’ perspectives on cancer care and artificial intelligence

BackgroundThe integration of Artificial Intelligence (AI) technology in cancer care has gained unprecedented global attention over the past few decades. This has impacted the way that cancer care is practiced and delivered across settings. The purpose of this study was to explore the perspectives and experiences of healthcare professionals (HCPs) on cancer treatment and the need for AI. This study is a part of the INCISIVE European Union H2020 project's development of user requirements, which aims to fully explore the potential of AI-based cancer imaging technologies.MethodsA mixed-methods research design was employed. HCPs participating in cancer care in the UK, Greece, Italy, Spain, Cyprus, and Serbia were first surveyed anonymously online. Twenty-seven HCPs then participated in semi-structured interviews. Appropriate statistical method was adopted to report the survey results by using SPSS. The interviews were audio recorded, verbatim transcribed, and then thematically analysed supported by NVIVO.ResultsThe survey drew responses from 95 HCPs. The occurrence of diagnostic delay was reported by 56% (n = 28/50) for breast cancer, 64% (n = 27/42) for lung cancer, 76% (n = 34/45) for colorectal cancer and 42% (n = 16/38) for prostate cancer. A proportion of participants reported the occurrence of false positives in the accuracy of the current imaging techniques used: 64% (n = 32/50) reported this for breast cancer, 60% (n = 25/42) for lung cancer, 51% (n = 23/45) for colorectal cancer and 45% (n = 17/38) for prostate cancer. All participants agreed that the use of technology would enhance the care pathway for cancer patients. Despite the positive perspectives toward AI, certain limitations were also recorded. The majority (73%) of respondents (n = 69/95) reported they had never utilised technology in the care pathway which necessitates the need for education and training in the qualitative finding; compared to 27% (n = 26/95) who had and were still using it. Most, 89% of respondents (n = 85/95) said they would be opened to providing AI-based services in the future to improve medical imaging for cancer care.Interviews with HCPs revealed lack of widespread preparedness for AI in oncology, several barriers to introducing AI, and a need for education and training. Provision of AI training, increasing public awareness of AI, using evidence-based technology, and developing AI based interventions that will not replace HCPs were some of the recommendations.ConclusionHCPs reported favourable opinions of AI-based cancer imaging technologies and noted a number of care pathway concerns where AI can be useful. For the future design and execution of the INCISIVE project and other comparable AI-based projects, the characteristics and recommendations offered in the current research can serve as a reference.

Using machine learning ensemble method for detection of energy theft in smart meters

AbstractElectricity theft is a primary concern for utility providers, as it leads to substantial financial losses. To address the issue, a novel extreme gradient boosting (XGBoost)‐based model utilizing the consumers’ electricity consumption patterns for analysis is proposed for electricity theft detection (ETD). To remove the imbalance in the real‐world electricity consumption dataset and ensure an even distribution of theft and non‐theft data instances, six different artificially created theft attacks were used. Moreover, the utilization of the XGBoost algorithm for classification, especially to identify malicious instances of electricity theft, yielded commendable accuracy rates and a minimal occurrence of false positives. The proposed model identifies electricity theft specific to the regions, utilizing electricity consumption parameters, and other variables as input features. The authors’ model outperformed existing benchmarks like k‐neural networks, light gradient boost, random forest, support vector machine, decision tree, and AdaBoost. The simulation results using the false attacks for balancing the dataset have improved the proposed model's performance, achieving a precision, recall, and F1‐score of 96%, 95%, and 95%, respectively. The results of the detection rate and the false positive rate (FPR) of the proposed XGBoost‐based detection model have achieved 96% and 3%, respectively.

Machine Learning-Based Lung Cancer Detection Using Multiview Image Registration and Fusion

The exact lung cancer identification is a critical problem that has attracted the researchers’ attention. The practice of multiview single image and segmentation has been widely used for the last 2 years to improve the identification of lung cancer disease. The utilization of machine learning (ML) and deep learning (DL) techniques can significantly expedite the process of cancer detection and stage classification, enabling researchers to study a larger number of patients in a shorter time frame and at a reduced cost applying the image segmentation approach herein, the multiresolution rigid registration mechanism is applied to enhance the segmentation further. Techniques like principle component averaging and discrete wavelet transform are verified for the image fusion development. To review the performance of the suggested technique, the image database resource initiative-based lungs image database consortium is tested in this paper which includes 4,682 computed tomography scan images of 61 patients with nodules sizes from 3 to 30 mm. According to the study finding, the outperformed results of our model are obtained in terms of feature mutual information, and peak signal-to-noise ratio, which were recorded at 0.80 and 19.25, respectively. Moreover, the detection and stages of cancer (STG-1, STG-2, STG-3, and STG-4) of lung nodules are also assessed by using the ResNet-18 convolutional neural network classifier. With only 1.8 FP/scan, the achieved accuracy and sensitivity for detection are 98.2% and 96.4%, respectively. The study’s findings show that our proposed strategy outperforms existing models significantly. Therefore, the proposed models have the potential to be implemented in clinical settings to provide support to doctors in the early diagnosis of cancer, while minimizing the occurrence of false positives in scans.

A Hierarchical Fusion SAR Image Change-Detection Method Based on HF-CRF Model

The mainstream methods for change detection in synthetic-aperture radar (SAR) images use difference images to define the initial change regions. However, methods can suffer from semantic collapse, which makes it difficult to determine semantic information about the changes. In this paper, we proposed a hierarchical fusion SAR image change-detection model based on hierarchical fusion conditional random field (HF-CRF). This model introduces multimodal difference images and constructs the fusion energy potential function using dynamic convolutional neural networks and sliding window entropy information. By using an iterative convergence process, the proposed method was able to accurately detect the change-detection regions. We designed a dynamic region convolutional semantic segmentation network with a two-branch structure (D-DRUNet) to accomplish feature fusion and the segmentation of multimodal difference images. The proposed network adopts a dual encoder–single decoder structure where the baseline is the UNet network that utilizes dynamic convolution kernels. D-DRUNet extracts multimodal difference features and completes semantic-level fusion. The Sobel operator is introduced to strengthen the multimodal difference-image boundary information and construct the dynamic fusion pairwise potential function, based on local boundary entropy. Finally, the final change result is stabilized by iterative convergence of the CRF energy potential function. Experimental results demonstrate that the proposed method outperforms existing methods in terms of the overall number of detection errors, and reduces the occurrence of false positives.

Reinforcement learning based distantly supervised biomedical named entity recognition

Data used by current Biomedical named entity recognition (BioNER) systems has mostly been manually labelled for supervision. However, it might be difficult to find large amounts of annotated data, especially in fields with a high level of specialization, such as biomedical, bioinformatics, and so on. When dictionaries and ontologies are available, which are domain-specific knowledge resources, automatically tagged distantly supervised biomedical training data can be developed. However, any such distantly supervised NER result is normally noisy. The prevalence of false positives and false negatives with this type of autonomously generated data is the main problem that directly affects efficiency. This research investigates distant supervision to detect false positive occurrences in BioNER task. A reinforcement learning technique is employed that is modelled as a graphical processing unit (GPU) accelerated Markov decision process (MDP) with a neural network policy. To deal with false negative cases, we employ a partial annotation conditional random field (CRF) technique. Results on two benchmark datasets show a cutting-edge methodology that can enhance the functionality of the neural NER system. It goes on to show how the proposed approach cuts down on human annotated data for BioNER tasks in Natural Language Processing (NLP).