False Errors Research Articles

Design of a fuzzy current‐sensor less maximum power point tracking algorithm for photovoltaic systems

In this study, a fuzzy current-sensor less maximum power point tracking (MPPT) algorithm is presented. The algorithm proposed to improve the performance and reduce the cost of photovoltaic systems. The proposed algorithm creates the variable step-size/variable frequency specification in the drift-free single input voltage sensor (SIVS) method designing a reliable and widely used fuzzy control algorithm. The proposed algorithm will improve the SIVS methods employing the advantages of the fuzzy controllers in the MPPT algorithms based on the real and false error areas analysis concept developed by Tousi et al. (2016). The proposed method will detect exact drift conditions and move to the optimal setpoint at the maximum speed. The method selects the correct direction, appropriate step-size, and suitable control frequency to increase dynamic efficiency against the drift condition. Also, the steady-state efficiency of the algorithm will increase because of high-frequency and low-power oscillation performances. The proposed method will be simulated and compared in Matlab/Simulink environment and tested in laboratory conditions on the designed solar charge controller.

Hybrid SVM-CNN Classification Technique for Human-Vehicle Targets in an Automotive LFMCW Radar.

Human–vehicle classification is an essential component to avoiding accidents in autonomous driving. The classification technique based on the automotive radar sensor has been paid more attention by related researchers, owing to its robustness to low-light conditions and severe weather. In the paper, we propose a hybrid support vector machine–convolutional neural network (SVM-CNN) approach to address the class-imbalance classification of vehicles and pedestrians with limited experimental radar data available. A two-stage scheme with the combination of feature-based SVM technique and deep learning-based CNN is employed. In the first stage, the modified SVM technique based on these distinct physical features is firstly used to recognize vehicles to effectively alleviate the imbalance ratio of vehicles to pedestrians in the data level. Then, the residual unclassified images will be used as inputs to the deep network for the subsequent classification, and we introduce a weighted false error function into deep network architectures to enhance the class-imbalance classification performance at the algorithm level. The proposed SVM-CNN approach takes full advantage of both the locations of underlying class in the entire Range-Doppler image and automatical local feature learning in the CNN with sliding filter bank to improve the classification performance. Experimental results demonstrate the superior performances of the proposed method with the score of and area under the curve (AUC) of the receiver operating characteristic (ROC) of over several state-of-the-art methods with limited experimental radar data available in a 77 GHz automotive radar.

Brain Tumor Classification and Segmentation Based on Morphological Operations using Image Processing Techniques

Brain Tumour is the solitary cause for the assassination of many individuals. A brain tumour is an accretion or widening of isolated cells in your brain. It can be identified by several tests like MRI (Magnetic Resonance Imaging), CT (Computer Tomography) scan along with several tests like biopsy, spinal tap etc. Classification and Segmentation activity take part a significant role in interpretation of brain tumours. In this paper the images should be taken in the form of jpeg format. The images are processed using data mining and machine learning classification methods. Previous research studies are intended as long as identifying brain tumours using dissimilar classification and segmentation approaches. The initiated system consists of certain process for recognition of the tumour. The first step is about Pre-processing and the next is about segmentation, Feature extraction is the third step in this process and Classification is used to detect the tumour. Morphological Operations are performed in this process based on the tumour size, shape and colour. Neural Network is used for classification along with -Support Vector Machine (SVM) classifiers are worn for structured recognition. Due to this we can reduce the inappropriate or false diagnosis error rate of brain tumour identification for the patients and also, we can get faster and accurate results.

Clustering with varying risks of false assignments in discrete latent variable model.

In clustering problems, to model the intrinsic structure of unlabeled data, the latent variable models are frequently used. These model-based clustering methods often provide a clustering rule minimizing the total false assignment error. However, in many clustering applications, it is desirable to treat false assignment errors for a certain cluster differently. In this paper, we introduce the false assignment rate for clustering and estimate it by using the extended likelihood approach. We propose VRclust, a novel clustering rule that controls various errors differently across clusters. Real data examples illustrate the usage of estimation of false assignment rate and a simulation study shows that error controls are consistent as the sample size increases.

Design and Analysis of Leakage-Induced False Error Tolerant Error Detecting Latch for Sub/Near-Threshold Applications

The digital designs operating in sub/near-threshold region are susceptible to timing errors due to the extreme impact of process, voltage, and temperature (PVT) variations. This paper proposes a new error-detecting latch (EDL) to mitigate the impact of PVT variations. The proposed EDL is a single-phase clocked design, which significantly reduces the clock power consumption of the design. The proposed EDL follows a merged and shared architecture of two latches along with an XNOR gate, which leads to a compact layout of EDL. The post-layout simulations in an industrial 28 nm CMOS technology node, show a minimum clock power savings of 31%, average power savings of 16%, and reduction in leakage power by 23% in comparison to the state-of-the-art EDLs at 0.4 V. The 10K rigorous Monte-Carlo simulations across the supply voltage range of 0.23 V - 0.8 V and clock frequency range of 1.6 MHz - 70 MHz shows that the proposed EDL is tolerant to leakage-induced false errors and glitches. SSEDL is robust to process variations even with minimum-sized transistors.

Spectrum Sensing in Cognitive Radio using Frequency Domain

An efficient bandwidth allocation and dynamic bandwidth access away from its previous limits is referred as cognitive radio (CR).The limited spectrum with inefficient usage requires the advances of dynamic spectrum access approach, where the secondary users are authorized to utilize the unused temporary licensed spectrum. For this reason it is essential to analyze the absence/presence of primary users for spectrum usage. So spectrum sensing is the main requirement and developed to sense the absence/ presence of a licensed user. This paper shows the design model of energy detection based spectrum sensing in frequency domain utilizing Binary Symmetric Channel (BSC) ,Additive white real Gaussian channel (AWGN), Rayleigh fading channel users for 16-Quadrature Amplitude Modulation(QAM) which is utilized for the wide band sensing applications at low Signal to noise Ratio(SNR) level to reduce the false error identification. The spectrum sensing techniques has least computational complexity. Simulink model for the energy detection based spectrum sensing using frequency domain in MATLAB 2014a.

Mitigating Coordinated Call Attacks On VoIP Networks Using Hidden Markov Model

Abstract This paper presents a 2-tier scheme for mitigating coordinated call attacks on VoIP networks. Call interaction pattern was considered using talk and salient periods in a VoIP call conversation. At the first-tier, Short Term Energy algorithm was used for call interaction feature extraction and at the second-tier Hidden Markov Model was used for caller legitimacy recognition. Data of VoIP call conversations were collated and analyzed to extract distinctive features in VoIP call interaction pattern to ascertain the legitimacy of a caller against coordinated call attacker. The performance metrics that was used are; False Error Rate (FER), Specificity, Detection Accuracy and Throughput. Several experiments were conducted to see how effective the mitigating scheme is, as the scheme acts as a proxy server to Session Initiation Protocol (SIP) server. The experiments show that; when the VoIP server is under coordinated call attack without a mitigating scheme only 15.2% of legitimate VoIP users had access to the VoIP network and out of which about half of the legitimate users had their calls dropped before completion, while with the 2-tier mitigating scheme, when the VoIP server is under coordinated call attacks over 90.3% legitimate VoIP callers had their calls through to completion

Real-manufacturing-oriented big data analysis and data value evaluation with domain knowledge

As one of the most popular topics currently, big data has played an important role in both academic research and practical applications. However, in the manufacturing industry, it is difficult to make full use of the research results for production optimization and/or management due to the low quality of real workshop data. Typical quality problems of real workshop data include the information match degree, missing recessive data, and false error identification. The conventional data analysis methods cannot handle most such issues because these methods fail to consider professional insights into and domain knowledge about the data. The main motivation of this paper is to explore methods for analyzing and evaluating big data with domain knowledge. For this purpose, real production data from a semiconductor manufacturing workshop are adopted as the data object. First, a series of data analysis techniques with domain knowledge are developed for diagnosing the imperfections. Then, corresponding data processing techniques with domain knowledge are proposed for solving those data quality problems according to specific flaws in the data. Furthermore, this paper proposes quantitative calculation methods of data value density to determine the extent to which data quality can be improved by the proposed data processing techniques. Case studies are conducted to demonstrate that data analysis and processing techniques with domain knowledge can effectively handle data quality problems of real workshop data in terms of the information match degree, missing recessive data, and false error identification. The work in this paper has the potential to be further extended and applied to other big data applications beyond the manufacturing industry.

Automatic Lung Cancer Detection using Sobel & Morphological Operations

Cancer is the most dangerous disease that may cause death and lung cancer is one of them which is more common among all. There are various imaging techniques through which organs can be scanned for diagnosis. Lung cancer is a disease that may be caused by unrestrained cell growth in lung. Lung cancer is the most common and most dangerous cancer. CT scan can obtain the lung images, but still it has been recognized manually. Manual lung cancer detection is a challenging task because false error rate may lead you to compromise with human’s life. There are lots of researches that has been done in this field but still failed to obtain high precision with minimal error rate. Here the system proposes automatic lung cancer detection using Sobel & Morphological operations that can acquire good precision along with cancer area detection. Sobel is a gradient edge detection technique through which absolute gradient magnitude is computed in the reference of 2D input lung image that is later dilated with morphological operator. The obtained result is liable to attain high precision with less false alarm rate.

Integration of a Commercial Barcode-Assisted Medication Dispensing System in a Teaching Hospital

A commercial barcode-assisted medication administration (BCMA) system was integrated to secure the medication process and particularly the dispensing stage by technicians and the administration stage with nurses. We aimed to assess the impact of this system on medication dispensing errors and barriers encountered during integration process. We conducted a controlled randomized study in a teaching hospital, during dispensing process at the pharmacy department. Four wards were randomized in the experimental group and control group, with two wards using the system during 3 days with dedicated pharmacy technicians. The system was a closed loop system without information return to the computerized physician order entry system. The two dedicated technicians had a 1-week training session. Observations were performed by one observer among the four potential observers previously trained. The main outcomes assessed were dispensing error rates and the identification of barriers encountered to expose lessons learned from this study. There was no difference between the dispensing error rate of the control and experimental groups (7.9% for both, p = 0.927). We identified 10 barriers to pharmacy barcode-assisted system technology deployment. They concerned technical (problems with semantic interoperability interfaces, bad user interface, false errors generated, lack of barcodes), structural (poor integration with local information technology), work force (short staff training period, insufficient workforce), and strategic issues (system performance problems, insufficient budget). This study highlights the difficulties encountered in integrating a commercial system in current hospital information systems. Several issues need to be taken into consideration before the integration of a commercial barcode-assisted system in a teaching hospital. In our experience, interoperability of this system with the electronic health record is the key for the success of this process with an entire closed loop system from prescription to administration. BCMA system at the dispensing process remains essential to purchase securing medication administration process.

Medicare fraud detection using neural networks

Access to affordable healthcare is a nationwide concern that impacts a large majority of the United States population. Medicare is a Federal Government healthcare program that provides affordable health insurance to the elderly population and individuals with select disabilities. Unfortunately, there is a significant amount of fraud, waste, and abuse within the Medicare system that costs taxpayers billions of dollars and puts beneficiaries’ health and welfare at risk. Previous work has shown that publicly available Medicare claims data can be leveraged to construct machine learning models capable of automating fraud detection, but challenges associated with class-imbalanced big data hinder performance. With a minority class size of 0.03% and an opportunity to improve existing results, we use the Medicare fraud detection task to compare six deep learning methods designed to address the class imbalance problem. Data-level techniques used in this study include random over-sampling (ROS), random under-sampling (RUS), and a hybrid ROS–RUS. The algorithm-level techniques evaluated include a cost-sensitive loss function, the Focal Loss, and the Mean False Error Loss. A range of class ratios are tested by varying sample rates and desirable class-wise performance is achieved by identifying optimal decision thresholds for each model. Neural networks are evaluated on a 20% holdout test set, and results are reported using the area under the receiver operating characteristic curve (AUC). Results show that ROS and ROS–RUS perform significantly better than baseline and algorithm-level methods with average AUC scores of 0.8505 and 0.8509, while ROS–RUS maximizes efficiency with a 4× speedup in training time. Plain RUS outperforms baseline methods with up to 30× improvements in training time, and all algorithm-level methods are found to produce more stable decision boundaries than baseline methods. Thresholding results suggest that the decision threshold always be optimized using a validation set, as we observe a strong linear relationship between the minority class size and the optimal threshold. To the best of our knowledge, this is the first study to compare multiple data-level and algorithm-level deep learning methods across a range of class distributions. Additional contributions include a unique analysis of the relationship between minority class size and optimal decision threshold and state-of-the-art performance on the given Medicare fraud detection task.

Evaluation of the detection and correction properties of the reference dictionary of the system for checking and correcting orthography

The models for evaluating the properties of the reference orthographic dictionary (ROD) of the spelling check and correction system are considered. RODs’ detecting properties are determined by the probability of not detecting the typical error and the probability of a false error notification. The task is formulated to optimize a ROD according to Pareto, a step by step algorithm is proposed for solving it, the results of the experimental evaluation of the algorithm’s effectiveness are given. RODs’ correcting properties are determined by the probabilities of the correct and erroneous correction of the typical errors. Models of their estimation are offered and simulation results are given for the selected dictionaries. It has been shown that ROD optimized for detecting properties also has better correcting properties. In general, the obtained results can be used as the basis for a tool for the comparative assessment, selection and improvement of the potential properties of a specific ROD for a given subject matter.

A Distributed Error and Anomaly Communication Architecture for Analog and Mixed-Signal Systems

With the increasing use of autonomous and mission critical systems, in field concurrent testing (i.e. testing while device is in operation) is becoming progressively more important. However, concurrent testing of analog and mixed-signal systems often needs to rely on anomaly indicators, because of the absence of clear 1/0 error conditions in analog circuits. In this paper we show how simultaneous monitoring of different but related anomalies across a mixed-signal system can lead to improved diagnosis of error conditions and rules out false errors. However, to achieve such kind of federated learning between different modules across the system, we need an architecture through which all modules within the system can communicate their anomaly information with minimum overhead. To that end, first, we develop a data structure (Weighted Alarm Collector - WAC) to efficiently encapsulate all the anomaly information in a system. Second, we show how WACs can be merged through hierarchies in a scalable manner. Based on this, we also develop an architecture to facilitate learning across different modules across hierarchies. The main benefits of this architecture are that a) it enables distributed error learning within a system and b) the architecture is scalable and can be built bottom up, which is compatible with current IP based modular design methodology. Finally we show some applications of this architecture on some common industry use-cases.

Computational Evaluation of Mixing Performance in 3-D Swirl-Generating Passive Micromixers

Computational Fluid Dynamics (CFD) tools are used to investigate fluid flow and scalar mixing in micromixers where low molecular diffusivities yield advection dominant transport. In these applications, achieving a numerical solution is challenging. Numerical procedures used to overcome these difficulties may cause misevaluation of the mixing process. Evaluation of the mixing performance of these devices without appropriate analysis of the contribution of numerical diffusion yields over estimation of mixing performance. In this study, two- and four-inlet swirl-generating micromixers are examined for different mesh density, flow and molecular diffusivity scenarios. It is shown that mesh densities need to be high enough to reveal numerical diffusion errors in scalar transport simulations. Two-inlet micromixer design was found to produce higher numerical diffusion. In both micromixer configurations, when cell Peclet numbers were around 50 and 100 for Reynolds numbers 240 and 120, the numerical diffusion effects were tolerable. However, when large cell Peclet number scenarios were tested, it was found that the molecular diffusivity of the fluid is completely masked by false diffusion errors.

Анализ корректности работы с памятью с использованием расширения теории символьных графов памяти предикатами над символьными значениями

Safety-critical systems require additional effort to comply with specifications. One of the required specification is correct memory usage. The article describes an efficient method for static verification against memory safety errors as a combination of Symbolic Memory Graphs and predicate abstraction on symbolic values used in graph. In this article, we introduce an extension of Symbolic Memory Graphs. In addition to symbolic values, the graph stores predicates over symbolic values, which allow to track the relationship between symbolic values in the graph. We also expand existing vertex types to support arbitrary abstract regions, which allow us to represent such dynamic data structures as lists and trees. One of the types of abstract regions is also the ODM region, which presents a special kind of on-demand memory that occurs when analyzing incomplete programs. For this memory, the size and structure of the contents are not known in advance, but it is believed that such memory can be operated safely. The method is implemented in CPAchecker tool. Practical usage is demonstrated on Linux kernel modules. The practical contribution of our work is to reduce false error messages by constructing more accurate abstractions using predicates over symbolic values.

Serial position-dependent false memory effects

ABSTRACTEvidence for false recognition within seconds of encoding suggests that semantic-associative influences are not restricted to long-term memory, consistent with unitary memory accounts but contrary to dual store models. The present study sought further relevant evidence using a modified free recall converging associates task where participants studied 12-item lists composed of 3 semantically distinct quartets (sublists) related to a separate, non-presented theme word (i.e., words 1–4/theme1, 5–8/theme2, and 9–12/theme3). This list construction permits assessment of false recall errors from each sublist, and, particularly, the primacy and recency sublists that have been linked to long- and short-term memory stores. Experiment 1 tested immediate free recall for items. Associative false memories were evident from all sublists, however, significantly less so from the recent sublist, which also showed the highest levels of veridical memory. By inserting a brief (3 s) distractor prior to recall, Experiment 2 selectively reduced veridical memory and increased false memory for the recent sublist while leaving the primacy sublist unaffected. These recall results converge with prior evidence indicating the immediacy of false recognition, and can be understood within a unitary framework where the differential availability of verbatim features and gist-based cues affect memory for primacy and recency sublists.

A ratiometric electrochemiluminescent biosensor for Con A detecting based on competition of dissolved oxygen

A novel ratiometric electrochemiluminescent (ECL) biosensor was designed for the detection of concanavalin A (Con A) based on two ECL emitters competing for the dissolved oxygen (O2). In this strategy, CdTe quantum dots (QDs) were used as the cathodic emitter and N-(aminobutyl)-N-(ethylisoluminol) (ABEI) was used as the anodic emitter. With the presence of dissolved O2 utilized as co-reactant, CdTe QDs showed an ECL emission at − 1.7 V, and ABEI showed an emission at + 0.6 V. Phenoxy-derivatized dextran (Dexp), as a recognition element, was immobilized by graphene oxide functionalized CdTe QDs (G-CdTe QDs) to recognize Con A, and further combine with Dexp, gold and platinum nanoparticles decorated ABEI (Dexp-Au-Pt-ABEI) to form a sandwich structure. With the increasing concentration of analyte Con A, the ECL signal of cathodic CdTe QDs decreased and the anodic response of ABEI increased, thus achieving a ratiometry detection of Con A with a wide linear range from 1.0 × 10−4 ng/mL to 10 ng/mL. The detection limit was low to 3.0 × 10−5 ng/mL. This proposed ratiometric ECL biosensor not only conquered the false errors caused by external interferences, but excluded the disability caused by exogenous co-reactant via the application of common dissolved O2. It also exhibited an excellent stability, selectivity and reproducibility.

A Scheme to Design Concurrent Error Detection Techniques for the Fast Fourier Transform Implemented in SRAM-Based FPGAs

Soft errors are an important issue for SRAM-based Field Programmable Gate Arrays (FPGAs), since they result in permanent alterations of the mapped circuit when they affect their configuration memory. Concurrent Error Detection (CED) techniques, such as Dual Modular Redundancy (DMR), are usually employed to detect errors that affect the performance of the circuit. When trying to detect errors produced on the complex Fast Fourier Transform (FFT), the Parseval Sum of Squares (SoS) is a widely used technique. In this paper, we present a scheme to implement CED techniques for the complex FFT implemented in SRAM-based FPGAs. These techniques perform checks based on the relationships existing between one or more of the inputs and the outputs of the algorithm. Three examples of these techniques are provided to further clarify how to construct them. These techniques, along with DMR and SoS, have been tested through fault injection. An analysis on their error detection capabilities shows that they achieve high detection rates with much less resource usage than DMR and SoS. In addition, the number of false error detections for these techniques is lower than that of SoS, which leads to less unnecessary reconfigurations of the device.

IMPROVEMENT OF THE PERFORMANCE OF FINGERPRINT VERIFICATION USING A COMBINATORIAL APPROACH

Fingerprint verification systems have attracted much attention in secure organizations; however, conventional methods still suffer from unconvincing recognition rate for noisy fingerprint images. To design a robust verification system, in this paper, wavelet and contourlet transforms (CTS) were suggested as efficient feature extraction techniques to elicit a coverall set of descriptive features to characterize fingerprint images. Contourlet coefficients capture the smooth contours of fingerprints while wavelet coefficients reveal its rough details. Due to the high dimensionality of the elicited features, across group variance (AGV), greedy overall relevancy (GOR) and Davis–Bouldin fast feature reduction (DB-FFR) methods were adopted to remove the redundant features. These features were applied to three different classifiers including Boosting Direct Linear Discriminant Analysis (BDLDA), Support Vector Machine (SVM) and Modified Nearest Neighbor (MNN). The proposed method along with state-of-the-art methods were evaluated, over the FVC2004 dataset, in terms of genuine acceptance rate (GAR), false acceptance rate (FAR) and equal error rate (EER). The features selected by AGV were the most significant ones and provided 95.12% GAR. Applying the selected features, by the GOR method, to the modified nearest neighbor, resulted in average EER of [Formula: see text]%, which outperformed the compared methods. The comparative results imply the statistical superiority ([Formula: see text]) of the proposed approach compared to the counterparts.

Behavioral and neurophysiological abnormalities during cued continuous performance tasks in patients with mild traumatic brain injury.

ObjectiveThis study's aim was to investigate the features and neural mechanisms of sustained attention in patients with mild traumatic brain injury (mTBI) by comparing and analyzing neuropsychological, behavioral, event‐related potentials, and event‐related desynchronization and synchronization between mTBI patients and healthy controls.MethodsTwenty mTBI patients with mTBI and 20 healthy controls underwent the Mini‐Mental State Examination (MMSE) and a cued continuous performance task (AX‐CPT). Neuropsychological, behavioral, and electroencephalogram (EEG) data were collected and analyzed.ResultsThere were significant differences between the mTBI group and the control group in their MMSE total scores, attention, and calculation, but there were no significant differences in orientation, memory, recall, and verbal scores. There were significant differences between the mTBI group and the control group in hitting the number, reaction time, and the number of errors of omission, but there were no significant differences in the number of false errors. The amplitude of Go‐N2 and Nogo‐N2 was significantly smaller for the mTBI group than that for the control group. The amplitude of Go‐P3 was significantly smaller for the mTBI group than that for the control group, but not for the amplitude of Nogo‐P3. The Go‐αERS were significantly less for the mTBI group than for the control group during the 0–200 ms after the stimulus onset. The Go‐αERD and Nogo‐αERD were significantly less for the mTBI group than for the control group during the 600–1,000 ms after the stimulus onset. The Go‐βERS were significantly less for the mTBI group than for the control group during the 200–400 ms after the stimulus onset. There were no significant differences in the Nogo‐αERS and Nogo‐βERD/ERS between the mTBI group and the control group.ConclusionPatients with mTBI exhibited impairments in sustained attention and conflict monitoring, while response inhibition may have been spared.