Bias Removal Research Articles

The effect of statistical normalization on network propagation scores.

Network diffusion and label propagation are fundamental tools in computational biology, with applications like gene-disease association, protein function prediction and module discovery. More recently, several publications have introduced a permutation analysis after the propagation process, due to concerns that network topology can bias diffusion scores. This opens the question of the statistical properties and the presence of bias of such diffusion processes in each of its applications. In this work, we characterized some common null models behind the permutation analysis and the statistical properties of the diffusion scores. We benchmarked seven diffusion scores on three case studies: synthetic signals on a yeast interactome, simulated differential gene expression on a protein-protein interaction network and prospective gene set prediction on another interaction network. For clarity, all the datasets were based on binary labels, but we also present theoretical results for quantitative labels. Diffusion scores starting from binary labels were affected by the label codification and exhibited a problem-dependent topological bias that could be removed by the statistical normalization. Parametric and non-parametric normalization addressed both points by being codification-independent and by equalizing the bias. We identified and quantified two sources of bias-mean value and variance-that yielded performance differences when normalizing the scores. We provided closed formulae for both and showed how the null covariance is related to the spectral properties of the graph. Despite none of the proposed scores systematically outperformed the others, normalization was preferred when the sought positive labels were not aligned with the bias. We conclude that the decision on bias removal should be problem and data-driven, i.e. based on a quantitative analysis of the bias and its relation to the positive entities. The code is publicly available at https://github.com/b2slab/diffuBench and the data underlying this article are available at https://github.com/b2slab/retroData. Supplementary data are available at Bioinformatics online.

Urbanisation’s contribution to climate warming in Great Britain

Urbanisation is changing the climate of the world we live in. In Great Britain (GB) 5.8% of the total land area is covered by artificial surfaces, increasing from 4.3% in 1975. Aside from associated loss of farmland, biodiversity and a range of ecosystem services, changing to urban form warms the Earth’s surface: the urban heat island (UHI) effect. Standard estimates of temperature changes do not account for urbanisation (i.e. use of rural-only stations or removal of urban bias in observations), meaning that anthropogenic modifications to the land surface may be causing the surface-level atmosphere to warm quicker than those estimates suggest. Using observations from a high-density urban monitoring network, we show that locally this warming (instantaneously) may be over 8 °C. Based on the relationships between UHI intensity, urban fraction and wind speed in this network, we create a statistical model and use it to estimate the current daily-mean urban warming across GB to be 0.04 °C [0.02 °C –0.06 °C]. Despite this climate contribution appearing small (94% of GB’s land cover for the time-being is still rural), we show that half of GB’s population currently live in areas with average daily-mean warming ∼0.4 °C. Under heatwave conditions our high estimates show 40% of GB’s population may experience over a 1 °C daily-mean UHI. Furthermore, simply due to urbanisation (1975–2014) we estimate GB is warming at a rate equivalent and in addition to 3.4% [1.9%–5.0%] of the observed surface-level warming calculated from background stations. In the fastest urbanising region, South East GB, we find that these warming rates are up to three times faster. The methodology is straightforward and can be readily extended to other countries or updated as future land cover data becomes available.

Analysis of Solar Power and Energy Variability Through Site Adaptation of Satellite Data With Quality Controlled Measured Solar Radiation Data

Abstract Satellite-based solar power data is becoming more and more important because of its continuous temporal and spatial availability. However, its reliability can be enhanced through quality control and calibration against ground-based measurement data. Here, a holistic methodology is employed for the adaptation of satellite-based data for estimating solar energy. For the purpose, high-quality ground-based measurement data and satellite-based datasets are assessed across 12 sites in three small islands located in the Indian Ocean. Initially, both datasets go through a rigorous quality control process. A quantitative analysis of irradiance and insolation data is then conducted. Eventually, site adaptation of satellite-based data is performed using bias removal technique and statistical analysis of datasets. A set of seven statistical performance indicators is used to support the assessment. Analysis of datasets shows that adaptation of peak values should be performed separately. Results showed that despite the small surface areas of the islands studied, a spatial variation of insolation can be depicted. A temporal variation of insolation is also noted with a peak in the summer and low insolation levels in winter. Peak irradiance values tend to exceed solar constant for all sites. Variations of peak irradiance can only be noticed in ground-based measurement data. While insolation levels are comparable in the summer season for all the sites, insolation levels in the winter season are higher in the sites with lower latitudes. Calibration factors for peak irradiance, monthly and annual average irradiance as well as yearly insolation are presented.

Relaxation of competing electromechanical couplings in murine artery

Piezoelectricity and pyroelectricity in biological tissues, which originate from oriented fibrous proteins with a polar axis, have long been suggested to play important roles in physiological functions. The possible manipulation of their polarity by external mechanisms, however, remains unsettled. We revisit this problem here using piezoresponse force microscopy (PFM) as the tool and the intima layer of murine artery as a model system. By carefully examining first and second harmonic piezoresponses at both selected points and through spatial mapping, we establish that electromechanical coupling probed by PFM is predominantly piezoelectric in the intima layer, while the quadratic effect makes only minor contributions. More importantly, we observe competition between the linear and quadratic effects after removal of DC biases applied to the sample surface, revealing not only interesting relaxation dynamics, but also highly asymmetric piezoresponse. Positive DC rotates dipoles in tropoelastin monomers away with reduced alignment, while negative DC aligns dipoles more leading to enhanced piezoresponse. The electric manipulation of biological polarity is thus demonstrated, with the relaxation time constant determined on the order of 0.1 s, much slower than classical ferroelectrics.

Improved prediction of fuel properties with near-infrared spectroscopy using a complementary sequential fusion of scatter correction techniques

Near-infrared (NIR) spectroscopy of fuels can suffer from scattering effects which may mask the signals corresponding to key analytes in the spectra. Therefore, scatter correction techniques are often used prior to any modelling so to remove scattering and improve predictive performances. However, different scatter correction techniques may carry complementary information so that, if jointly used, both model stability and performances could be improved. A solution to that is the fusion of complementary information from differently scatter corrected data. In the present work, the use of a preprocessing fusion approach called sequential preprocessing through orthogonalization (SPORT) is demonstrated for predicting key quality parameters in diesel fuels. In particular, the possibility of predicting four different key properties, i.e., boiling point (°C), density (g/mL), aromatic mass (%) and viscosity (cSt), was considered. As a comparison, standard partial least-squares (PLS) regression modelling was performed on data pretreated by SNV and 2nd derivative (which is a widely used preprocessing combination). The results showed that the SPORT models, based on the fusion of scatter correction techniques, outperformed the standard PLS models in the prediction of all the four properties, suggesting that selection and use of a single scatter correction technique is often not sufficient. Up to complete bias removal with 50% reduction in prediction error was obtained. The R2P was increased by up to 8%. The sequential scatter fusion approach (SPORT) is not limited to NIR data but can be applied to any other spectral data where a preprocessing optimization step is required.

Removal of bias introduced by considering calendar or rainfall day as 24‐hr period in daily minimum temperature series: Results from ACMANT approach

AbstractThe automation of weather stations influences the climate observations in various ways, among them the hour of reading and reporting daily temperature extremes is usually altered. This alteration causes a systematic change, referred to as terminus change signal (TCS), in the recorded daily minimum temperatures (Tmin), in comparison with the earlier records. The TCS present in the Tmin data of the Meteorological Service of Catalonia is quantified accurately from the dates of the automation and hourly temperature observations. The TCS is very small for the annual means, but it has a characteristic, irregular‐shaped annual cycle with up to 0.3°C decrease of mean monthly Tmin from spring to autumn, and a change of the opposite sign in the winter months. The study explores how biases of the same seasonal behaviour as the TCS in Catalonia can be removed with statistical homogenization when the archived data and metadata are insufficient for the exact calculation of the bias properties. ACMANT, one of the most modern and most accurate homogenization methods is used on various test datasets with the TCS embedded. The test datasets are developed from two source datasets, one is a selected network of the observed Catalan Tmin series, while the other is a simulated Swedish Tmin dataset from the European project INDECIS. The test datasets include the TCS with varied parameterizations in order to obtain a more general picture about the skill of the statistical homogenization. While the signal‐to‐noise ratio and mean annual bias are varied by parameterization, the shape of the annual cycle of the TCS is fixed. The results show that ACMANT is notably more accurate in removing mean annual biases than mean monthly biases. The skill of removing a mean monthly bias widely varies and is influenced by several factors identified by the performed tests.

Importance of old bulls: leaders and followers in collective movements of all-male groups in African savannah elephants (Loxodonta africana)

In long-lived social species, older individuals can provide fitness benefits to their groupmates through the imparting of ecological knowledge. Research in this area has largely focused on females in matrilineal societies where, for example, older female African savannah elephants (Loxodonta africana) are most effective at making decisions crucial to herd survival, and old post-reproductive female resident killer whales (Orcinus orca) lead collective movements in hunting grounds. In contrast, little is known about the role of older males as leaders in long-lived social species. By analysing leadership patterns of all-male African savannah elephant traveling groups along elephant pathways in Makgadikgadi Pans National Park, Botswana, we found that the oldest males were more likely to lead collective movements. Our results challenge the assumption that older male elephants are redundant in the population and raise concerns over the biased removal of old bulls that currently occurs in both legal trophy hunting and illegal poaching. Selective harvesting of older males could have detrimental effects on the wider elephant society through loss of leaders crucial to younger male navigation in unknown, risky environments.

Teaching further education students the effects of naive realism, to support social development and mitigate classroom conflict

Aim(s)Naive realism is the false conviction that one perceives and responds to the world objectively or ‘as it is’ rather than subjectively through the lens of personal motives, experiences and expectations. The teaching of naïve realism to adults has evidenced increased perspective taking. The current research aimed to evaluate if this approach would be effective in supporting adolescents experiencing classroom conflict.MethodA critical realist approach was employed. Participants were a class of fifteen 16–19-year-old further education (FE) students and related academic staff. Students attended two one-hour lessons on naïve realism. Thematic analysis was conducted on pre- and post-teaching focus group data.FindingsPre-intervention, students identified feeling that others were acting belligerently and presenting inaccurate narratives, yet defended their own similar behaviour as routine. Unexpectedly, students also reported a conscious and purposeful use of naïve realism. Post-intervention, students and staff reported: substantial conflict resolution; improved abilities to perspective take, empathise and self-regulate; reduced fear of negative peer evaluation and anxiety; improved learning atmosphere and applications to the wider world. Novel findings include: the benefits of teaching as a whole-class approach; and how bias removal was insufficient to facilitate perspective taking in certain circumstances.LimitationsThe subjective and interpretative nature of the research methodology, small, all-female sample, and potential for social desirability bias, limit the transferability of findings.ConclusionsTeaching adolescents about the effects of naïve realism may have the potential to support wellbeing, relationships, attainment and transition to adulthood.

Wi-Fi RTT Ranging Performance Characterization and Positioning System Design

The aim of this research is to implement a precise Wi-Fi indoor positioning system (IPS) or localization system based upon the IEEE 802.11mc fine-timing measurement (FTM) scheme also known as the Wi-Fi round trip time (RTT) ranging technique, where ranging refers to a sub-process of positioning that determines the distance between a transmitter and receiver. Our system and its algorithms were implemented using a COTS (Commercial-Off-The-Shelf) smartphone and Wi-Fi access points. Experiments were conducted in several real-life indoor environments. This paper presents the detailed Wi-Fi RTT ranging performance of these devices in different system configurations and characterizes the systematic biases and noise model to improve the ranging accuracy. A novel three-step-positioning method is proposed to overcome the issues of no or multiple intersect points in trilateration due to ranging errors to improve positioning accuracy. This consists of the following: 1) systematic bias determination and removal; 2) clustering-based trilateration (CbT) supported by weighted concentric circle generation (WCCG), namely CbT & WCCG; 3) positioning result and trajectory optimization using a Kalman filter. As a result, the evaluation experiments gave a position accuracy of ±1.2 m in 2D static positioning and ±1.3 m for dynamic motion tracking. Also, our CbT & WCCG method demonstrate good tolerance against ranging errors. Moreover, the computational cost and positioning accuracy of CbT & WCCG methods are compared with least square (LS) and recursive least square (RLS) methods and the accuracy standard deviation of our algorithm is the closest to the Cramer–Rao bound (CRB).

Meteorological trends over Satluj River Basin in Indian Himalaya under climate change scenarios

Temperature and precipitation distributions depend on variable topography and heterogeneous landuse/landcover in the Indian Himalayan Region (IHR). It imparts a major concern for hydrological, glaciological modelling, dam structure assessment, etc. Thus, there is an inherent requirement of robust information for climate impact studies over the topographical variable and landuse heterogenous region in the Indian Himalayan Region (IHR). In particular, the importance of bias corrections become critically important over Himalayan river basins, in which model outputs with the corresponding in-situ observations are used for improving the model distribution. These improved details in present and future, as well, are important to carry out the climate change impact studies at basin scale for hydrological, glaciological, climatological studies, etc. Thus, in the present study firstly, model fields are bias corrected with the corresponding in-situ observations. And then trends in these bias corrected data is compared with the corresponding in-situ observations. These assessment of present and future changes in temperature and precipitation over Satluj River Basin (SRB) located in the western Himalayas is illustrated. Model fields are considered from a Regional climate model (REMO) from Coordinated Regional Downscaling Experiment-South Asia (CORDEX-SA) in three Representative Concentration Pathways (RCPs), i.e., 2.6, 4.5 and 8.5 W/m2. These projections are bias corrected using distributed quantile mapping. The precipitation (temperature) bias correction is performed using the distributed quantile mapping on the gamma (normal) distribution. The standard trend statistics is applied for quantitative assessment. A good capture of bias correction in temperature and precipitation is illustrated. Efficient bias removal is depicted in cumulative distribution curve (CDF) at individual station. Trend analysis shows that highest rate of precipitation decrement at low altitude station (Kasol) with the rate of −6.362 mm/year in RCP 8.5. Over the SRB highest rate of temperature increment is seen at highest altitude station (Kaza) with the rate of 0.084°C/year in RCP 8.5. On an average, fall in precipitation and increase in temperature with >99% confidence level in RCP 8.5 is seen. In addition, intensity lowers in other lower RCPs. The study sums up with the efficacy of CORDEX-SA REMO model in capturing present and future change in temperature and precipitation over the SRB in western Himalayas using the bias correction.

Performance analysis of bias correction techniques in brain MR images

Bias field is a smooth intensity variation, which emanate during the process of image procurement. Bias removal is an essential prerequisite while incorporating computer assisted diagnosis. Several bias correction algorithms are proposed till date. This paper scrutinizes the prominent bias correction algorithms, LEMS, MICO, BCFCM and N3. These investigations carried over the substantial amount of T1 and T2 weighted brain MR images with different bias spectrum from Brain Web website. Algorithms efficiency are analyzed in spectral wise, slice wise, and type wise. Based upon the performance indicators coefficient of variation and coefficient of joint variation, the algorithms are assessed and ranked. The result concludes that which algorithm exterminates the bias field, presents in brain MR images in an efficient manner.

Oxygen extraction fraction mapping with multi-parametric quantitative BOLD MRI: Reduced transverse relaxation bias using 3D-GraSE imaging

Magnetic resonance imaging (MRI)-based quantification of the blood-oxygenation-level-dependent (BOLD) effect allows oxygen extraction fraction (OEF) mapping. The multi-parametric quantitative BOLD (mq-BOLD) technique facilitates relative OEF (rOEF) measurements with whole brain coverage in clinically applicable scan times. Mq-BOLD requires three separate scans of cerebral blood volume and transverse relaxation rates measured by gradient-echo (1/T2∗) and spin-echo (1/T2). Although the current method is of clinical merit in patients with stroke, glioma and internal carotid artery stenosis (ICAS), there are relaxation measurement artefacts that impede the sensitivity of mq-BOLD and artificially elevate reported rOEF values.We posited that T2-related biases caused by slice refocusing imperfections during rapid 2D-GraSE (Gradient and Spin Echo) imaging can be reduced by applying 3D-GraSE imaging sequences, because the latter requires no slice selective pulses. The removal of T2-related biases would decrease overestimated rOEF values measured by mq-BOLD. We characterized effects of T2-related bias in mq-BOLD by comparing the initially employed 2D-GraSE and two proposed 3D-GraSE sequences to multiple single spin-echo reference measurements, both in vitro and in vivo. A phantom and 25 participants, including young and elderly healthy controls as well as ICAS-patients, were scanned. We additionally proposed a procedure to reliably identify and exclude artefact affected voxels. In the phantom, 3D-GraSE derived T2 values had 57% lower deviation from the reference. For in vivo scans, the formerly overestimated rOEF was reduced by −27% (p ​< ​0.001). We obtained rOEF ​= ​0.51, which is much closer to literature values from positron emission tomography (PET) measurements. Furthermore, increased sensitivity to a focal rOEF elevation in an ICAS-patient was demonstrated.In summary, the application of 3D-GraSE improves the mq-BOLD-based rOEF quantification while maintaining clinically feasible scan times. Thus, mq-BOLD with non-slice selective T2 imaging is highly promising to improve clinical diagnostics of cerebrovascular diseases such as ICAS.

An updated long‐term homogenized daily temperature data set for Australia

AbstractA new version of the long‐term Australian temperature data set, known as ACORN‐SAT (Australian Climate Observations Reference Network—Surface Air Temperature), has been developed. ACORN‐SAT includes homogenized daily maximum and minimum temperature data from 112 locations across Australia, encompassing the period from 1910 to the present, with 60 of the locations having data for the full 1910–2018 period. Homogenization is achieved using a percentile‐matching methodology with a number of improvements beyond practices used in previous versions, including more effective detection and removal of potentially inhomogeneous reference stations and an enhanced breakpoint detection methodology. Explicit corrections have also been introduced for a change in instrument screen size, whilst an assessment has found that the transition from manual to automatic instruments and changes in effective response time of automatic instruments have had a negligible impact on the data. Adjustments associated with documented site moves from in‐town to out‐of‐town locations are predominantly negative, particularly for minimum temperature, with other adjustments showing no strong bias towards either positive or negative values. The new data set shows slightly stronger warming (0.12°C per decade in mean temperature over the 1910–2016 period) than either the previous ACORN‐SAT version (0.10°C) or the unhomogenized gridded data (0.08°C), primarily due to more effective treatment of systematic moves of sites out of towns and the removal of a rounding bias in the version 1 methodology.

Precipitation Trends Over Mainland China From 1961–2016 After Removal of Measurement Biases

AbstractFollowing the removal of measurement biases from daily observed gauge records for the period 1961–2016 at 2,250 national stations over mainland China, the precipitation trends and causes were analyzed. Measured annual precipitation trends have been overestimated to varying degrees in most regions, with the exceptions of Northwest China and northern Tibet. On average, the annual precipitation trend over mainland China decreased from 3.98 to 2.04 mm/10 yr after bias adjustment. The dramatically precipitation‐increasing (wetting) zones in Northeast China and the lower reaches of the Yangtze River were found to be disappearing at a faster rate than what we observed. Meanwhile, the extent of markedly precipitation‐decreasing (drying) zones in the Yellow River Basin shrank more slowly than what was indicated by measured records. Adjusted annual light rain amounts exhibited a weaker increasing trend over mainland China; conversely, adjusted annual rainstorm amounts displayed stronger increasing trends than reported previously. The changes of trace precipitation days, precipitation days, mean wind speed on precipitation days, and observed precipitation amount comprehensively affected the trends of bias‐removed precipitation. The declines in measured wind speed and precipitation amount resulted in accelerated reductions in adjusted precipitation over Southwest and North China. In Southeast China, the increasing trend of bias‐adjusted precipitation was smaller than that indicated by original data due to the reduction in mean wind speed when precipitation occurred. In Northwest China and northern Tibet, however, the increases in measured precipitation amount and precipitation days mutually led to larger increasing trends in the adjusted data.

Extreme Learning Machine-Based State Reconstruction for Automatic Attack Filtering in Cyber Physical Power System

Successful detection of false data injection attacks (FDIAs) and removal of state bias due to FDIAs are essential for ensuring secure power grids operation and control. This article first extends the approximate dc model of FDIA to a more general ac model that can handle both traditional and synchronized measurements. To automatically filter out the established FDIAs, we propose a state reconstruction scheme consisting of a contaminated state separation method, an enhanced bad data identification approach and a state recovery algorithm. In this scheme, a classifier is developed by aggregating a series of extreme learning machines (ELMs) to detect anomaly states caused by FDIAs. Gaussian random distribution and Latin hypercube sampling are adopted to initialize the input weights of base ELMs, which can provide more diversities to enhance the ensemble performance. Then, to identify the exact locations of the compromised measurements, a state forecasting-based bad data identification approach is proposed by exploiting the consistency between the forecasted and the received measurements. Finally, an effective state recovery algorithm applies quasi-Newton method and Armijo line search to address the possible system unobservable problem due to the removal of attacked measurements. Numerical tests on serval IEEE standard test systems verify the efficiency of the proposed FDIA model and state reconstruction scheme.

Efficiency of propensity score adjustment and calibration on the estimation from non-probabilistic online surveys

One of the main sources of inaccuracy in modern survey techniques, such as online and smart- phone surveys, is the absence of an adequate sampling frame that could provide a probabilistic sampling. This kind of data collection leads to the presence of high amounts of bias in final es- timates of the survey, specially if the estimated variables (also known as target variables) have some influence on the decision of the respondent to participate in the survey. Various correction techniques, such as calibration and propensity score adjustment or PSA, can be applied to re- move the bias. This study attempts to analyse the efficiency of correction techniques in multiple situations, applying a combination of propensity score adjustment and calibration on both types of variables (correlated and not correlated with the missing data mechanism) and testing the use of a reference survey to get the population totals for calibration variables. The study was performed using a simulation of a fictitious population of pot ential voters and a real volunteer survey aimed to a population for which a complete census was available. Results showed that PSA combined with calibration results in a bias removal considerably larger when compared with calibration with no prior adjustment. Results also showed that using population totals from the estimates of a reference survey instead of the available population data does not make a difference in estimates accuracy, although it can contribute to slightly increment the variance of the estimator.

Toward a high-quality pan-genome landscape of Bacillus subtilis by removal of confounding strains.

Pan-genome analysis is widely used to study the evolution and genetic diversity of species, particularly in bacteria. However, the impact of strain selection on the outcome of pan-genome analysis is poorly understood. Furthermore, a standard protocol to ensure high-quality pan-genome results is lacking. In this study, we carried out a series of pan-genome analyses of different strain sets of Bacillus subtilis to understand the impact of various strains on the performance and output quality of pan-genome analyses. Consequently, we found that the results obtained by pan-genome analyses of B. subtilis can be influenced by the inclusion of incorrectly classified Bacillus subspecies strains, phylogenetically distinct strains, engineered genome-reduced strains, chimeric strains, strains with a large number of unique genes or a large proportion of pseudogenes, and multiple clonal strains. Since the presence of these confounding strains can seriously affect the quality and true landscape of the pan-genome, we should remove these deviations in the process of pan-genome analyses. Our study provides new insights into the removal of biases from confounding strains in pan-genome analyses at the beginning of data processing, which enables the achievement of a closer representation of a high-quality pan-genome landscape of B. subtilis that better reflects the performance and credibility of the B. subtilis pan-genome. This procedure could be added as an important quality control step in pan-genome analyses for improving the efficiency of analyses, and ultimately contributing to a better understanding of genome function, evolution and genome-reduction strategies for B. subtilis in the future.

A symmetric D flip-flop based PUF with improved uniqueness

Physically unclonable functions (PUF) emerged as security primitives that generate high entropy, temper resilient bits for security applications. However, the implementation area budget limits their use in lightweight applications such as IoT, RFID, and biomedical applications. In the form of SRAM or D flip-flop, intrinsic PUFs are abundantly available in almost all of the designs. Being an integral part of the design, they can be used with compromised performance. In this work, to address the usage of intrinsic PUF, a D flip-flop based lightweight PUF is proposed. The proposed architecture is implemented on 40 nm CMOS technology. The simulation results show that it offers a uniqueness of 0.502 and the worst-case reliability of 95.89% at high temperature 125 °C and 97.89% at a supply voltage of 1.2 V. To evaluate the performance of various PUF architectures, A novel term, the uniqueness-to-reliability ratio, is proposed. When compared to the conventional D flip-flop, it offers 4.491 times more uniqueness and 127.74 times more uniqueness-to-reliability ratio with the same layout area. Since it uses the symmetrical structure, unlike other architectures, the proposed architecture does not require any post-processing schemes for bias removal, which further saves the silicon area. To verify the functional correctness of the simulation results, an FPGA implementation of the conventional and proposed D Flip-flop is also presented.

A Simple Model for Subject Behavior in Subjective Experiments

In a subjective experiment to evaluate the perceptual audiovisual quality of multimedia and television services, raw opinion scores offered by subjects are often noisy and unreliable. Recommendations such as ITU-R BT.500, ITU-T P.910 and ITU-T P.913 standardize post-processing procedures to clean up the raw opinion scores, using techniques such as subject outlier rejection and bias removal. In this paper, we analyze the prior standardized techniques to demonstrate their weaknesses. As an alternative, we propose a simple model to account for two of the most dominant behaviors of subject inaccuracy: bias (aka systematic error) and inconsistency (aka random error). We further show that this model can also effectively deal with inattentive subjects that give random scores. We propose to use maximum likelihood estimation (MLE) to jointly estimate the model parameters, and present two numeric solvers: the first based on the Newton-Raphson method, and the second based on alternating projection. We show that the second solver can be considered as a generalization of the subject bias removal procedure in ITU-T P.913. We compare the proposed methods with the standardized techniques using real datasets and synthetic simulations, and demonstrate that the proposed methods have advantages in better model-data fit, tighter confidence intervals, better robustness against subject outliers, shorter runtime, the absence of hard coded parameters and thresholds, and auxiliary information on test subjects. The source code for this work is open-sourced at https://github.com/Netflix/sureal.

Distributed Scatterer Interferometry With the Refinement of Spatiotemporal Coherence

The state-of-the-art techniques have demonstrated that coherence error degrades the performance of synthetic aperture radar (SAR) interferometry (InSAR) for distributed scatterers (DSs). This article aims at fully evaluating the influence of coherence error on DS InSAR time-series analysis. In particular, we present a methodology to increase the estimation accuracy of DS interferometry, with emphasis on spatiotemporal coherence refinement. The motive behind this is that bias removal and variance mitigation of sample coherence matrix impose optimum weighting for estimating phase series and geophysical parameters of interest, whereas maximization of temporal coherence in a reference network can avoid spatial error propagation during the least-squares adjustment. Rather than developing independent processing chains, we integrate this method into SqueeSAR technique and simultaneously take the advantage of StaMPS into consideration. Using simulation and real data over southwestern China, comprehensive comparisons before and after spatiotemporal coherence refinement are performed over various coherence scenarios. The results tested from different phase and displacement rate estimators validate the effectiveness of the presented method.