Small Error Rate Research Articles

Quality of routine health data collected by health workers using smartphone at primary health care in Ethiopia

BackgroundMobile phone based applications are considered by many as potentially useful for addressing challenges and improving the quality of data collection in developing countries. Yet very little evidence is available supporting or refuting the potential and widely perceived benefits on the use of electronic forms on smartphones for routine patient data collection by health workers at primary health care facilities. MethodsA facility based cross sectional study using a structured paper checklist was prepared to assess the completeness and accuracy of 408 electronic records completed and submitted to a central database server using electronic forms on smartphones by 25 health workers. The 408 electronic records were selected randomly out of a total of 1772 maternal health records submitted by the health workers to the central database over a period of six months. Descriptive frequencies and percentages of data completeness and error rates were calculated. ResultsWhen compared to paper records, the use of electronic forms significantly improved data completeness by 209 (8%) entries. Of a total 2622 entries checked for completeness, 2602 (99.2%) electronic record entries were complete, while 2393 (91.3%) paper record entries were complete. A very small percentage of error rates, which was easily identifiable, occurred in both electronic and paper forms although the error rate in the electronic records was more than double that of paper records (2.8% vs. 1.1%). More than half of entry errors in the electronic records related to entering a text value. ConclusionsWith minimal training, supervision, and no incentives, health care workers were able to use electronic forms for patient assessment and routine data collection appropriately and accurately with a very small error rate. Minimising the number of questions requiring text responses in electronic forms would be helpful in minimizing data errors.

Generic power attacks on RSA with CRT and exponent blinding: new results

Schindler and Itoh (Applied cryptography and network security-ACNS 2011. Lecture Notes in Computer Science, vol 6715. Springer, Berlin, pp 73–90, 2011) and Schindler and Wiemers (J Cryptogr Eng 4:213–236, 2014. doi:10.1007/s13389-014-0081-y) treat generic power attacks on RSA implementations (with CRT/without CRT) and on ECC implementations (scalar multiplication with the long-term key), which apply exponent blinding, resp., scalar blinding, as algorithmic countermeasure against side-channel attacks. In Schindler and Itoh (2011) and Schindler and Wiemers (2014), it is assumed that an adversary has guessed the blinded exponent bits/the blinded scalar bits independently for all power traces and for all bit positions, and each bit guess is false with probability \(\epsilon _b>0\). Three main types of attacks and several variants thereof were introduced and analysed in Schindler and Itoh (2011) and Schindler and Wiemers (2014). The attacks on RSA with CRT are the least efficient since the attacker has no information on \(\phi (p)\). In this paper, we introduce two new attack algorithms on RSA with CRT, which improve the attack efficiency considerably. In particular, attacks on blinding factors of length \(R=64\) have definitely become practical, and for small error rates \(\epsilon _b\) even \(R=96\) may be overcome.

FIJI Macro 3D ART VeSElecT: 3D Automated Reconstruction Tool for Vesicle Structures of Electron Tomograms

Automatic image reconstruction is critical to cope with steadily increasing data from advanced microscopy. We describe here the Fiji macro 3D ART VeSElecT which we developed to study synaptic vesicles in electron tomograms. We apply this tool to quantify vesicle properties (i) in embryonic Danio rerio 4 and 8 days past fertilization (dpf) and (ii) to compare Caenorhabditis elegans N2 neuromuscular junctions (NMJ) wild-type and its septin mutant (unc-59(e261)). We demonstrate development-specific and mutant-specific changes in synaptic vesicle pools in both models. We confirm the functionality of our macro by applying our 3D ART VeSElecT on zebrafish NMJ showing smaller vesicles in 8 dpf embryos then 4 dpf, which was validated by manual reconstruction of the vesicle pool. Furthermore, we analyze the impact of C. elegans septin mutant unc-59(e261) on vesicle pool formation and vesicle size. Automated vesicle registration and characterization was implemented in Fiji as two macros (registration and measurement). This flexible arrangement allows in particular reducing false positives by an optional manual revision step. Preprocessing and contrast enhancement work on image-stacks of 1nm/pixel in x and y direction. Semi-automated cell selection was integrated. 3D ART VeSElecT removes interfering components, detects vesicles by 3D segmentation and calculates vesicle volume and diameter (spherical approximation, inner/outer diameter). Results are collected in color using the RoiManager plugin including the possibility of manual removal of non-matching confounder vesicles. Detailed evaluation considered performance (detected vesicles) and specificity (true vesicles) as well as precision and recall. We furthermore show gain in segmentation and morphological filtering compared to learning based methods and a large time gain compared to manual segmentation. 3D ART VeSElecT shows small error rates and its speed gain can be up to 68 times faster in comparison to manual annotation. Both automatic and semi-automatic modes are explained including a tutorial.

Error rates and variation between observers are reduced with the use of photographic matching software for capture-recapture studies

Photographic capture-mark-recapture (CMR) permits individual recognition whilst avoiding many of the concerns involved with marking animals. However, the construction of capture histories from photographs is a time-consuming process. Furthermore, matching accuracy is determined based on subjective judgements of the person carrying out the matching, which can lead to errors in the resulting datasets – particularly in long-term projects where multiple observers match images. We asked 63 volunteers to carry out two photographic-matching exercises using a database of known individuals of the yellow-bellied toad (Bombina variegata). From these exercises, we quantified the matching accuracy of volunteers in terms of false-acceptance and false-rejection rates. Not only were error rates greatly reduced with the use of photographic-matching software, but variation in error rates among volunteers was also lowered. Furthermore, the use of matching software led to substantial increases in matching speeds and an 87% reduction in the false-rejection rate. As even small error rates have the potential to bias CMR analyses, these results suggest that computer software could substantially reduce errors in CMR datasets. The time-savings and reduction in variance among observers suggest that such methods could be particularly beneficial in long-term CMR projects where a large number of images may be matched by multiple observers.

Combined Use of Rear Touch Gestures and Facial Feature Detection to Achieve Single-Handed Navigation of Mobile Devices

To improve single-handed operation of mobile devices, rear touch panel operation to control commands and facial feature detection to control cursor position are proposed. Operational control is achieved through finger chord gestures on a rear touch panel, and nose movement is used to control cursor movement. Zooming is achieved by detecting the apparent distance between the left and right eyes in conjunction with a finger chord gesture. We have evaluated movement time, error rates, and the throughputs of these techniques in comparison with the conventional single-handed front touch panel thumb operations using Fitts's law. Experiments have been conducted to evaluate two operation modes, selection and zooming, in the form of reciprocal 1-D pointing tasks with 12 participants. For the target selection task, the proposed technique achieved 12% (0.26 s) shorter movement time and 4.7% smaller error rate than the conventional method on average. Especially, for long distance targets, the performance of the conventional method became remarkably inferior due to the limit of reach of the thumb, whereas the proposed technique achieved much less deterioration and obtained expected performance because the cursor could reach anywhere on the display. For the target size adjustment task, the proposed technique achieved 9% (0.22 s) shorter movement time than the conventional method and obtained a comparable error rate of less than 4%. Consequently, we could demonstrate the techniques that make single-handed select and zoom operations available anywhere on a large-sized tablet device with no blockage of the display.

Advantages of Array-Based Technologies for Pre-Emptive Pharmacogenomics Testing.

As recognised by the National Institutes of Health (NIH) Precision Medicine Initiative (PMI), microarray technology currently provides a rapid, inexpensive means of identifying large numbers of known genomic variants or gene transcripts in experimental and clinical settings. However new generation sequencing techniques are now being introduced in many clinical genetic contexts, particularly where novel mutations are involved. While these methods can be valuable for screening a restricted set of genes for known or novel mutations, implementation of whole genome sequencing in clinical practice continues to present challenges. Even very accurate high-throughput methods with small error rates can generate large numbers of false negative or false positive errors due to the high numbers of simultaneous readings. Additional validation is likely to be required for safe use of any such methods in clinical settings. Custom-designed arrays can offer advantages for screening for common, known mutations and, in this context, may currently be better suited for accredited, quality-controlled clinical genetic screening services, as illustrated by their successful application in several large-scale pre-emptive pharmacogenomics programs now underway. Excessive, inappropriate use of next-generation sequencing may waste scarce research funds and other resources. Microarrays presently remain the technology of choice in applications that require fast, cost-effective genome-wide screening of variants of known importance, particularly for large sample sizes. This commentary considers some of the applications where microarrays continue to offer advantages over next-generation sequencing technologies.

Random subspace method with class separability weighting

AbstractThe random subspace method (RSM) is one of the ensemble learning algorithms widely used in pattern classification applications. RSM has the advantages of small error rate and improved noise insensitivity due to ensemble construction of the base‐learners. However, randomness may cause a reduction of the final ensemble decision performance because of contributions of classifiers trained by subsets with low class separability. In this study, we present a new and improved version of the RSM by introducing a weighting factor into the combination phase. One of the class separability criteria, J3, is used as a weighting factor to improve the classification performance and eliminate the drawbacks of the standard RSM algorithm. The randomly selected subsets are quantified by computing their J3 measure to determine voting weights in the model combination phase, assigning lower voting weight to classifiers trained by subsets with poor class separability. Two models are presented including J3‐weighted RSM and optimized J3 weighted RSM. In J3 weighted RSM, computed weighting values are directly multiplied by class assignment posteriors, whereas in optimized J3 weighted RSM, computed weighting values are optimized by a pattern search algorithm before multiplying by posteriors. Both models are shown to provide better error rates at lower subset dimensionality.

Too good to be true: when overwhelming evidence fails to convince

Is it possible for a large sequence of measurements or observations, which support a hypothesis, to counterintuitively decrease our confidence? Can unanimous support be too good to be true? The assumption of independence is often made in good faith; however, rarely is consideration given to whether a systemic failure has occurred. Taking this into account can cause certainty in a hypothesis to decrease as the evidence for it becomes apparently stronger. We perform a probabilistic Bayesian analysis of this effect with examples based on (i) archaeological evidence, (ii) weighing of legal evidence and (iii) cryptographic primality testing. In this paper, we investigate the effects of small error rates in a set of measurements or observations. We find that even with very low systemic failure rates, high confidence is surprisingly difficult to achieve; in particular, we find that certain analyses of cryptographically important numerical tests are highly optimistic, underestimating their false-negative rate by as much as a factor of 280.

IMPLEMENTATION LEARNING AND FORGETTING CURVE TO PREDICT NEEDS AND DECREASE OF LABORS PERFORMANCE AFTER BREAK

A company needs to implement production planning to minimize time and cost. Forecasting and scheduling are two methods which should be conducted in production planning. By implementing the learning and forgetting curve methods, the labor needs as well as the decrease of labors performance after break can be predicted. Firstly, various learning curve models are presented, then each model was analyzed one by one so that the model with the smallest error rate could be determined. A case study conducted in the learning curve model is presented with data derived from the production floor. The four main purposes of this study were to calculate the percentage of each station learning curve, learning and forgetting curves in the company, minimum initial cost, and predict the number of employees needed for the lowest in the number of the work station company. The results in the percentage achieved for the learning curve is 91.47%, the gluing station 78.46%, variation sewing station 98.10%, thumb sewing station 88.17%, omo connect sewing station 89.65%, machine sewing station 87.33%, omo folding sewing station 85.42, rubber tide sewing station 92.51%, sewing station tide studs 72.37%, omo tape sewing station 61.74%, and vilcro sewing station 75.89%, respectively. By analyzing the percentage of each station learning curve, a comparison between the highest and lowest percentage learning curve on the company was made. Thus, it is known that omo tape sewing station needs another operator as the additional labor. The percentage of the forgetting curve is 91.59%. Through a search conducted on the cumulative hours of the productive company, the initial cost of production can be minimized to 15.600 Indonesian rupiah.

Impacts of Species Misidentification on Species Distribution Modeling with Presence-Only Data

Spatial records of species are commonly misidentified, which can change the predicted distribution of a species obtained from a species distribution model (SDM). Experiments were undertaken to predict the distribution of real and simulated species using MaxEnt and presence-only data “contaminated” with varying rates of misidentification error. Additionally, the difference between the niche of the target and contaminating species was varied. The results show that species misidentification errors may act to contract or expand the predicted distribution of a species while shifting the predicted distribution towards that of the contaminating species. Furthermore the magnitude of the effects was positively related to the ecological distance between the species’ niches and the size of the error rates. Critically, the magnitude of the effects was substantial even when using small error rates, smaller than common average rates reported in the literature, which may go unnoticed while using a standard evaluation method, such as the area under the receiver operating characteristic curve. Finally, the effects outlined were shown to impact negatively on practical applications that use SDMs to identify priority areas, commonly selected for various purposes such as management. The results highlight that species misidentification should not be neglected in species distribution modeling.

Accurate interrogation of FCGR3A rs396991 in European and Asian populations using a widely available TaqMan genotyping method.

A polymorphism in the receptor for the Fc region of IgG, Fc γ-receptor IIIa (FcγRIIIa, FCGR3A rs396991), has been inconsistently shown in the literature to have an effect on response to monoclonal antibody therapy in several indications. The rs396991 (T/G) polymorphism leads to an F176V substitution and increased affinity for IgG. This variant has proven difficult to genotype accurately, primarily because of extensive homology between the FCGR3A and FCGR3B genes. We have shown that rs396991 can be genotyped by PCR amplification, followed by direct Sanger sequencing of the product, without coamplification of FCGR3B, and that the rs396991 TaqMan assay (C__25815666_10) agrees with Sanger sequencing results in 100% of European and Asian samples tested, but it has a small error rate in African and American populations. C__25815666_10 is therefore suitable to interrogate rs396991 in studies involving Europeans and Asians; however for other populations, the default genotyping method should be PCR followed by Sanger sequencing.

Adaptive management of multimodal biometrics fusion using ant colony optimization

This paper presents a new approach for the adaptive management of multimodal biometrics to meet a wide range of application dependent adaptive security requirements. In this work, ant colony optimization (ACO) is employed for the selection of key parameters like decision threshold and fusion rule, to ensure the optimal performance in meeting varying security requirements during the deployment of multimodal biometrics systems. Particle swarm optimization (PSO) has been widely utilized for the optimal selection of these parameters in the earlier attempts in the literature [Veeramachaneni et al., 2005] and [Kumar et al., 2010]. However, in PSO these parameters are computed in continuous domain while they are assumed to be better represented as discrete variables [Kumar et al., 2010]. This paper therefore proposes the use of ACO, in which discrete biometric verification parameters are computed to ensure the optimal performance from the multimodal biometrics system. The proposed ACO based framework is also extended to the pattern classification approach where fuzzy binary decision tree (FBDT) is utilized for two-class biometrics verification. The experimental results are presented on true multimodal systems from various publicly available databases; IITD databases of palmprint and iris, XM2VTS database of speech and faces, and the NIST BSSR1 databases of faces and fingerprint images. Our experimental results presented in this paper suggest that (i) ACO based approach is capable of operating on significantly small error rates in comparison to the widely employed PSO for automated selection of biometrics fusion rules/parameters, (ii) the score-level fusion yields better performance with lower error rate in comparison to the decision level fusion, and finally (iii) the FBDT based classification approach delivers considerably superior performance for the adaptive biometrics verification.

Thin Ice Detection in the Barents and Kara Seas With AMSR-E and SSMIS Radiometer Data

We have studied thin ice detection using Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) and Special Sensor Microwave Imager/Sounder radiometer data acquired over the Barents and Kara Seas during three winters (November-April) in 2008-2011. Moderate Resolution Imaging Spectroradiometer-based ice thickness charts were used as reference data. Thin ice detection was studied using polarization and spectral gradient ratios (PR and GR) calculated from the 36/37 and 89/91 GHz radiometer data. Thresholds for thin ice detection and maximum thicknesses for the detected thin ice (hT) were determined, as were error rates for misdetections. The results for different 1-D PR and GR parameters led to the conclusion that the AMSR-E PR36 and H-polarized GR8936 would be the best parameters for a 2-D classifier. We adopted the linear discrimination analysis (LDA) as a statistical tool. Thin ice areas with hT of 30 cm could be separated from thicker ice fields with approximately 20% error level. In our large data set, the estimation of thin ice thickness was not possible with reasonable accuracy due to the large scatter between ice thickness and the PR and GR signatures. This is likely due to a large data set, besides thin ice in polynyas also thin ice in the marginal ice zone and thin ice from freeze-up period. The optimal LDA parameters in the classifier and hT depended on the daily mean air temperature ((T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">am</sub> )). We could not yet parameterize the classifier optimally according to (T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">am</sub> ), but the constructed classifier worked rather robustly as indicated by the relative small error rate variation between the three analyzed winters.

Low relative error in consumer-grade GPS units make them ideal for measuring small-scale animal movement patterns.

Consumer-grade GPS units are a staple of modern field ecology, but the relatively large error radii reported by manufacturers (up to 10 m) ostensibly precludes their utility in measuring fine-scale movement of small animals such as insects. Here we demonstrate that for data collected at fine spatio-temporal scales, these devices can produce exceptionally accurate data on step-length and movement patterns of small animals. With an understanding of the properties of GPS error and how it arises, it is possible, using a simple field protocol, to use consumer grade GPS units to collect step-length data for the movement of small animals that introduces a median error as small as 11 cm. These small error rates were measured in controlled observations of real butterfly movement. Similar conclusions were reached using a ground-truth test track prepared with a field tape and compass and subsequently measured 20 times using the same methodology as the butterfly tracking. Median error in the ground-truth track was slightly higher than the field data, mostly between 20 and 30 cm, but even for the smallest ground-truth step (70 cm), this is still a signal-to-noise ratio of 3:1, and for steps of 3 m or more, the ratio is greater than 10:1. Such small errors relative to the movements being measured make these inexpensive units useful for measuring insect and other small animal movements on small to intermediate scales with budgets orders of magnitude lower than survey-grade units used in past studies. As an additional advantage, these units are simpler to operate, and insect or other small animal trackways can be collected more quickly than either survey-grade units or more traditional ruler/gird approaches.

Noise Suppression Phenomenon in Fiber Raman Amplifier

Abstract The problem of small bit error rate is actual in fiber Raman amplifiers. Our experimental data on the amplified spontaneous emission (ASE) with backward pumping as a base of optical noise evaluation are presented in this paper. Raman gain spectrum for useful Stokes low-powered signal is calculated and compared with obtained spectrum of non-coherent ASE noise. It is shown that output optical signal has better gain than the amplified spontaneous emission that leads to appreciable growth of output optical signal-to-noise ratio.

Shim Coils Design for Halbach Magnet Based on Target Field Method

A modified target field method is presented to the construction of shim and gradient coils for Halbach-type magnet featuring cylindrical magnet and transversal magnetic field direction. With Biot–Savart law, magnetic field is deduced from current density distribution, and two solutions corresponding to relations between spherical harmonic functions degree m and current density degree k, namely $$m=k\pm 1$$ are derived. Then combined with magnetic field distribution of target points, current density coefficient $$U_{\rm q}$$ is adopted to calculate stream function, based on which the coils' structure can be obtained. An appropriate objective function integrating dissipated energy and relative error is utilized to further improve the design. Examples of X2–Y2 and Y coils indicate that solution corresponding to $$m=k-1$$ exhibits good agreement with theory in terms of obtaining lower dissipated energy and smaller error rate, suggesting that the algorithm provides an effective, preferable approach for the design of Halbach shim coils.

On the Capacity of Slotted Aloha With Ancillary Channels

In this letter, we focus on a legacy system operated with slotted-Aloha and complemented by a redundancy channel where nodes transmit replicas of their packets. The number of replicas follows a probability distribution, and successive interference cancelation is applied across channels. Leaning on the theory of codes on graphs and algebraic tools, we prove that the system can provide arbitrarily small error rate up to a certain load, beyond which packet losses have to be undergone with finite probability. Tight upper bounds on capacity are derived for both regions, characterizing the achievable performance as a function of the deployed ancillary resources. Simulation results for moderate MAC frame length are also provided.

Time- and Space-Efficient Sliding Window Top-k Query Processing

A sliding window top-k ( top-k/w ) query monitors incoming data stream objects within a sliding window of size w to identify the k highest-ranked objects with respect to a given scoring function over time. Processing of such queries is challenging because, even when an object is not a top-k/w object at the time when it enters the processing system, it might become one in the future. Thus a set of potential top-k/w objects has to be stored in memory while its size should be minimized to efficiently cope with high data streaming rates. Existing approaches typically store top-k/w and candidate sliding window objects in a k-skyband over a two-dimensional score-time space. However, due to continuous changes of the k-skyband, its maintenance is quite costly. Probabilistic k-skyband is a novel data structure storing data stream objects from a sliding window with significant probability to become top-k/w objects in future. Continuous probabilistic k-skyband maintenance offers considerably improved runtime performance compared to k-skyband maintenance, especially for large values of k , at the expense of a small and controllable error rate. We propose two possible probabilistic k-skyband usages: ( i ) When it is used to process all sliding window objects, the resulting top-k/w algorithm is approximate and adequate for processing random-order data streams. ( ii ) When probabilistic k-skyband is used to process only a subset of most recent sliding window objects, it can improve the runtime performance of continuous k-skyband maintenance, resulting in a novel exact top-k/w algorithm. Our experimental evaluation systematically compares different top-k/w processing algorithms and shows that while competing algorithms offer either time efficiency at the expanse of space efficiency or vice-versa, our algorithms based on the probabilistic k-skyband are both time and space efficient.

Fast community detection by SCORE

Consider a network where the nodes split into $K$ different communities. The community labels for the nodes are unknown and it is of major interest to estimate them (i.e., community detection). Degree Corrected Block Model (DCBM) is a popular network model. How to detect communities with the DCBM is an interesting problem, where the main challenge lies in the degree heterogeneity. We propose a new approach to community detection which we call the Spectral Clustering On Ratios-of-Eigenvectors (SCORE). Compared to classical spectral methods, the main innovation is to use the entry-wise ratios between the first leading eigenvector and each of the other leading eigenvectors for clustering. Let $A$ be the adjacency matrix of the network. We first obtain the $K$ leading eigenvectors of $A$, say, $\hat{\eta}_1,\ldots,\hat{\eta}_K$, and let $\hat{R}$ be the $n\times (K-1)$ matrix such that $\hat{R}(i,k)=\hat{\eta}_{k+1}(i)/\hat{\eta}_1(i)$, $1\leq i\leq n$, $1\leq k\leq K-1$. We then use $\hat{R}$ for clustering by applying the $k$-means method. The central surprise is, the effect of degree heterogeneity is largely ancillary, and can be effectively removed by taking entry-wise ratios between $\hat{\eta}_{k+1}$ and $\hat{\eta}_1$, $1\leq k\leq K-1$. The method is successfully applied to the web blogs data and the karate club data, with error rates of $58/1222$ and $1/34$, respectively. These results are more satisfactory than those by the classical spectral methods. Additionally, compared to modularity methods, SCORE is easier to implement, computationally faster, and also has smaller error rates. We develop a theoretic framework where we show that under mild conditions, the SCORE stably yields consistent community detection. In the core of the analysis is the recent development on Random Matrix Theory (RMT), where the matrix-form Bernstein inequality is especially helpful.

Two-Term Edgeworth Expansions for the Classes of U- and V-statistics

Much effort has been devoted to deriving Edgeworth expansions for various classes of statistics that are asymptotically normally distributed, with derivations tailored to the individual structure of each class. Expansions with smaller error rates are needed for more accurate statistical inference. Two such Edgeworth expansions are derived analytically in this paper. One is a two-term expansion for the standardized U-statistic of order m, m ⩾ 3, with an error rate o(n− 1). The other is an expansion with the same error rate for the distribution of the standardized V-statistic of the same order. In deriving the Edgeworth expansion, we made use of the close connection between the V- and U-statistics, which permits to first derive the needed expansion for the related U-statistic, then extend it to the V-statistic, taking into consideration the estimation of all difference terms between the two statistics.