Search

Measurement invariance holds when a latent construct is measured in the same way across different levels of background variables (continuous or categorical) while controlling for the true value of that construct. Using Monte Carlo simulation, this paper compares the multiple indicators, multiple causes (MIMIC) model and MIMIC-interaction to a novel use of alignment optimization (AO) for detecting measurement noninvariance when the violator is a continuous variable. Results showed that MIMIC and MIMIC-interaction in sequential likelihood ratio tests and Wald tests with a Bonferroni correction provided a good balance between identifying invariant and noninvariant (linear violations) items when n ≥ 500 in terms of classification accuracy (CA). AO (CA ≥ .86) was as competitive as MIMIC and MIMIC-interaction to linear invariance violations but was far better under nonlinear quadratic violations when n ≥ 1,000 (i.e., 100 per group for 10 groups). [ABSTRACT FROM AUTHOR]

This paper examines different geographically weighted (GW) approaches for calculating spatially distributed measures of accuracy / uncertainty, consolidating current approaches and proposing 2 new ones. GW frameworks use a moving window or kernel to extract and weight data subsets, from which local (ie spatially distributed) statistics or metrics are calculated. A validation dataset with hard and soft classifications is used to illustrate the approaches. It contains observed field survey data (also commonly derived from higher resolution imagery), and predicted data from a fuzzy c-means classification. The hard classes were used to estimate spatially distributed measures of overall, user's and producer's accuracies in two ways. First, by conceptualising them as probabilities to be estimated from generalised linear regression models (GLMs), extended into Geographically Weighted GLMs. Second, by constructing local GW correspondence matrices and then calculating local accuracy measures from these. The soft classes were used to calculate per-class measures of fuzzy certainty from the absolute difference between predicted and observed fuzzy memberships. Then, a novel fuzzy certainty logic is proposed and used to create fuzzy confusion matrices and per-class measures of fuzzy omission and commission error, supporting measures of fuzzy user's and producer's certainties. These were extended to the GW case to generate spatially distributed measures. Finally, the soft classifications were conceptualised as compositional data and measures of difference were estimated using Aitchison distances. In each case, the local hard and soft accuracy and certainty measures were interpolated over a 1 km grid to estimate accuracy surfaces. The context for this review is the increasing operational use of training and validation data, often with high numbers of records, containing both hard and soft classes. The data and R code used to undertake all the analyses in this paper are provided, supporting more nuanced analyses of such data. [ABSTRACT FROM AUTHOR]

The presence of noise results in the decrease of the data quality, and thus affects seriously the learning performance. In this paper, we investigate theoretically and empirically how adjusting the temperature value of the softmax function to control the prediction uncertainty can help improve the model robustness to noise. Moreover, we design a modified loss function, where a regularisation term based on the uncertainty of distinguishing those non-ground truth labels is incorporated in the loss function to simultaneously improve the training accuracy and the model robustness to noise. This study theoretically and experimentally validates that the temperature parameter of the softmax function significantly impacts the classification accuracy and the proposed loss function is effective in improving the performance in the presence of noise. The study provides some practical guidelines for improving classifier performance through clearly expressing and understanding the relationship between uncertainty and noise robustness. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new approach to zone identification which is based on considering features with high semantic richness. Out of the scenarios of selecting features for identifying a zone based on classifying the sentences in a text, we came to notice that in the scenario where specialized names belonging to a text's domain and mode of the verbs together with reduced versions of conventional features, including history, are taken into account, an accuracy rate of 61% (resp. 81%) is obtained which is higher than that belonging to Liakata's (resp. Fisas's) approach. Also, to have a genuine comparison, both Liakata's and Fisas's corpora are used in our experiments. Such accuracy is obtained at the place where less computational cost for extracting the features was decreased. In order to improve the accuracy of zone identification, a decision-level fusion process based on combining the results of separate classifiers, was considered. With regard to this, two fusion techniques of 'majority voting' and 'average of probabilities' were used. Experimentations show the fact that ensemble of 'Logistic Regression', 'Support Vector Machine' and 'Neural Network' as classifiers yields the best performance. Also 'majority voting' was shown to perform a bit better than 'average of probabilities'. [ABSTRACT FROM AUTHOR]

There has been tremendous growth in research on measurement invariance over the past two decades. However, given that psychological tests are commonly used for making classification decisions such as personnel selections or diagnoses, surprisingly, there has been little research on how noninvariance impacts classification accuracy. Millsap and Kwok previously proposed a selection accuracy framework for that purpose, which has been recently extended to categorical data. Their framework, however, only deals with classification using a unidimensional test. In contrast, classification in practice usually involves multidimensional tests (e.g., personality) or multiple tests, with different weights assigned to each dimension. In the current paper, we extend Millsap and Kwok's framework for examining the impact of noninvariance to a multidimensional test on classification. We also provide an R script for the proposed method and illustrate it with a personnel selection example using data from a published report featuring a five-factor personality inventory. [ABSTRACT FROM AUTHOR]

The accuracy of the results of the national tests in English, mathematics and science taken by 11-year olds in England has been a matter of much debate since their introduction in 1994, with estimates of the proportion of students incorrectly classified varying from 10 to 30%. Using live data from the 2009 and 2010 administration of the national tests, this paper uses a number of models, drawing on both classical and modern test theories, to explore the relationship between test reliability, and the extent of misclassification when a student’s test score is reported in terms of one of a small number of discrete levels of achievement. The results indicate that across the two cohorts (2009 and 2010) and six models, the averages of classification accuracy of the tests were about 85%, 90% and 87% in English, mathematics and science, respectively. Moreover, the different models yielded very similar results; the standard deviations of the values of classification accuracy generated were 1.9% for English, 1.0% for mathematics and 1.3% for science. [ABSTRACT FROM PUBLISHER]

Motivation of a subject, who is associated with the data acquisition of brain computer interface (BCI) experiment, is a very crucial parameter for executing a successful BCI application. This paper proposes a novel method to present the distribution of motivation of a subject during a BCI experiment. The proposed method was successfully applied to the BCI Competition 2003 Data Set III and the BCI Competition 2005 Data Set I using fast Fourier transform-based band power features with a linear discriminant analysis classifier. The results show that not only the motivation of the subject dramatically changes during the trial but also using highly motivated time segments provides 7.86% and 2.00% improvement in the classification accuracy of the BCI Competition 2003 Data Set III and the BCI Competition 2005 Data Set I, respectively. [ABSTRACT FROM AUTHOR]

The traditional classification is based on the assumption that distribution of indicator variable X in one class is homogeneous. However, when data in one class comes from heterogeneous distribution, the likelihood ratio of two classes is not unique. In this paper, we construct the classification via an ambiguity criterion for the case of distribution heterogeneity of X in a single class. The separated historical data in each situation are used to estimate the thresholds respectively. The final boundary is chosen as the maximum and minimum thresholds from all situations. Our approach obtains the minimum ambiguity with a high classification accuracy allowing for a precise decision. In addition, nonparametric estimation of the classification region and theoretical properties are derived. Simulation study and real data analysis are reported to demonstrate the effectiveness of our method. [ABSTRACT FROM AUTHOR]

This paper mainly presents an analysis of the effects of the impeller blade geometry on the performance of a turbo pneumatic separator based on the CFD simulations by using Fluent 14.5.0. The simulation results indicate that the air vortex can be eliminated by using the positive bending arc blades with a remarkably small impacting angle. Furthermore, the optimum installation angle range for the positive bending arc blades was studied. The material classification experiment results show that for the positive bending arc blades with the optimum installation angles, the fishhook depth is decreased, the classification accuracy is increased, and the particle size distribution of fine powders is narrowed, especially under the design condition, compared to the straight blades. This suggestion of impeller blade geometry for the turbo pneumatic separator is the important basis for designing this type of separators and improving its classification performances. [ABSTRACT FROM AUTHOR]

This paper proposes an efficient classification technique that utilises colour-converted hybrid segmentation algorithm with multi-class-support vector machine classifier for an automated diagnosis and classification of brain magnetic resonance (MR) images. The texture, colour and shape features have been extracted and these features are used to classify MR brain images into three categories, namely, normal, benign and malignant. This paper utilises colour-converted segmentation algorithm which uses hybrid particle swarm optimisation plus K-means clustering technique. Feature selection is performed by a sequential floating forward selection. The method proposed makes use of classification technique based on multi-class support vector machine to classify MR brain images. The accuracy of the proposed system has been verified and found that accuracy of above 97% can be achieved. The proposed system can provide best classification performance with high accuracy, low error rate and lower computational effort. [ABSTRACT FROM AUTHOR]

Land cover classification is a vital application area in the satellite image processing domain. Texture is a useful feature in land cover classification. In this paper, we propose a distributed texture-based land cover classification algorithm using Hidden Markov Model (HMM). Here, HMM is used for texture-based classification of remotely sensed images. Furthermore, to enhance the performance, data-intensive remotely sensed image is segmented and distributed into parallel sessions. Experiments were conducted on IRS P6 LISS-IV data, and the results were evaluated based on the confusion matrix, classification accuracy, and Kappa statistics. These results indicate that the proposed algorithm achieves a classification accuracy of 88.75%. [ABSTRACT FROM AUTHOR]

Combining data of several tests or markers for the classification of patients according to their health status for assigning better treatments is a major issue in the study of diseases such as cancer. In order to tackle this problem, several approaches have been proposed in the literature. In this paper, a step-by-step algorithm for estimating the parameters of a linear classifier that combines several measures is considered. The optimization criterion is to maximize the area under the receiver operating characteristic curve. The algorithm is applied to different simulated data sets and its performance is evaluated. Finally, the method is illustrated with a prostate cancer staging database. [ABSTRACT FROM AUTHOR]

One of the concerns regarding the tax collection system is incorrect case selection. Manual selection of audit cases by auditors (whose role is to detect individual cases of tax non-compliance) based on their expert knowledge of the taxpayers' behaviour, cannot uncover all patterns of non-compliant behaviour hidden in historical data. In addition, random selection of audit cases is not focused on the highest risks. In other words, manual selection has a high opportunity cost if it is used as the sole selection method. Computational intelligence provides methods, techniques and tools, which have been taught to automatically make accurate income tax predictions based on past observations. The data were retrieved from the real time environmental situation. Application of computational intelligence methods proved to be efficient in learning a classification algorithm to classify compliant and non-compliant taxpayers. The new algorithm was evaluated and validated in empirical tests on the same dataset. Although this algorithm had the same performance measurement as Bagging, it outperformed the other existing multiple classifiers in terms of performance. This illustrates an automated system that replicates the investigative operation of human tax risk auditors. [ABSTRACT FROM AUTHOR]

The accuracy of performance validity tests (PVTs) with culturally diverse populations has increasingly been questioned. High false positive rates have been found in some PVTs in culturally and linguistically diverse individuals within the U.S. and internationally. No study to date has investigated the accuracy of PVTs with Chinese-speaking immigrants (CSI) in the U.S. The current study aimed to evaluate two PVTs, the Test of Memory Malingering (TOMM) and Dot Counting Test (DCT), to determine their accuracy in a community sample of CSI with limited English proficiency. These two measures were used in a simulation design, contrasting 52 participants who were instructed to respond honestly to 22 participants instructed to feign incompetency to stand trial. Results demonstrated the scores of TOMM Trial 1 and Trial 2 were effective in classifying honest responders from simulators, whereas the DCT E-score did not differentiate the groups better than chance. However, false positive rates for the TOMM Trial 1, Trial 2, and the DCT E-score were relatively low. Only one honest responder (1.9%) was classified as exerting insufficient effort in TOMM Trial 1 and DCT E-score, and the TOMM Trial 2 did not misclassify any honest responders. Implications and cautionary statements are provided and discussed. [ABSTRACT FROM AUTHOR]

Smart health-care is the recent technology, which ensures the early diagnosis and prevention to the patients in the remote area. Particularly, for the serious illness, like cardiac problems, brain abnormalities require immediate attention. Moreover, owing to the commencement of the smart systems, the data from distributed sources is bulky, which imposes the complexity to handle and degrades the diagnosis accuracy. Hence, this research proposes an innovative distributed learning based classification model named federated learning dependent Tawny flamingo-based deep CNN classifier for the disease classification, which handles the clinical data from the distributed sources and ensures the disease classification with good accuracy. In this research, a tawny flamingo-based deep CNN is proposed to detect the Alzheimer's disease, where the parameter of the classifier is tuned by the tawny flamingo algorithm. The tawny flamingo optimization reveals the maximal accuracy of 98.252% for K-fold and 97.995% for training percentage dependent analysis. [ABSTRACT FROM AUTHOR]

In vegetation remote sensing, classification accuracy cannot be fixed, due to seasonal variations in spectral reflectance characteristics. This study aims to clarify the seasonal variability of classification accuracy by forest-cover type. In particular, this paper describes seasonal variability by each band or band combinations. The study area is located in the vicinities of Hisayama and Sasaguri in Fukuoka Prefecture, Japan. Natural broadleaved, coniferous plantation, and bamboo forests were studied. Supervised classification was applied to six SPOT/HRV images taken in 1997. Kappa analysis was applied to assess the classification accuracy and compare any two error matrices. The results revealed that some single band or two-band combinations were as accurate as, or more accurate than, the full band (all three bands). The disadvantages of using a full band were especially apparent in the season with high classification accuracy. This study indicates that using all given bands does not necessarily result in the highest classification accuracy. This study also suggests that band selection within the scope of forest type and seasonal variability can contribute to better forest-cover-type classifications. [ABSTRACT FROM AUTHOR]

Identifying an optimal cutoff value for a continuous biomarker is often useful for medical applications. For binary outcome, commonly used cutoff finding criteria include Youden's index, classification accuracy, and the Euclidean distance to the upper left corner on the ROC curve. We extend these three criteria to accommodate censored survival time that subjected to competing risks. We provide various definitions of time-dependent true positive rate and false positive rate and estimate those quantities using nonparametric methods. In simulation studies, the Euclidean distance to the upper left corner on the ROC curve shows the best overall performance. [ABSTRACT FROM AUTHOR]

This study was conducted to analyze potential item parameter drift (IPD) impact on person ability estimates and classification accuracy when drift affects an examinee subgroup. Using a series of simulations, three factors were manipulated: (a) percentage of IPD items in the CAT exam, (b) percentage of examinees affected by IPD, and (c) item pool targeting. IPD impact on ability estimation was evaluated using bias and root mean square error (RMSE). A multivariate general linear model treating three factors as fixed effects and bias and RMSE as outcome variables are used to investigate both main and interaction effects. The impact on classification accuracy was examined based on number and percentages of misclassifications, their significance, and rank order correlations. The findings revealed that IPD exposed to a sub-group of examinees can affect classification accuracy of those examinees substantially, but IPD impact on average ability estimation was small. The study provides useful information to schools, states, and countries planning to implement or are currently implementing CAT as part of their assessments by emphasizing IPD in subgroups may violate fairness and consistency principles of CAT applications. [ABSTRACT FROM AUTHOR]

Reliable and up-to-date urban land cover information is valuable in urban planning and policy development. Due to the increasing demand for reliable land cover information there has been a growing need for robust methods and datasets to improve the classification accuracy from remotely sensed imagery. This study sought to assess the potential of the newly launched Landsat 8 sensor’s thermal bands and derived vegetation indices in improving land cover classification in a complex urban landscape using the support vector machine classifier. This study compared the individual and combined performance of Landsat 8’s reflective, thermal bands and vegetation indices in classifying urban land use-land cover. The integration of Landsat 8 reflective bands, derived vegetation indices and thermal bands overall produced significantly higher accuracy classification results than using traditional bands as standalone (i.e. overall, user and producer accuracies). An overall accuracy above 89.33% and a kappa index of 0.86, significantly higher than the one obtained with the use of the traditional reflective bands as a standalone data-set and other analysis stages. On average, the results also indicate high producer and user accuracies (i.e. above 80%) for most of the classes with a McNemar’sZscore of 9.00 at 95% confidence interval showing significant improvement compared with classification using reflective bands as standalone. Overall, the results of this study indicate that the integration of the Landsat 8’s OLI and TIR data presents an invaluable potential for accurate and robust land cover classification in a complex urban landscape, especially in areas where the availability of high resolution datasets remains a challenge. [ABSTRACT FROM AUTHOR]

Recent work has shown that the Lasso-based regularization is very useful for estimating the high-dimensional inverse covariance matrix. A particularly useful scheme is based on penalizing the ℓ1 norm of the off-diagonal elements to encourage sparsity. We embed this type of regularization into high-dimensional classification. A two-stage estimation procedure is proposed which first recovers structural zeros of the inverse covariance matrix and then enforces block sparsity by moving non-zeros closer to the main diagonal. We show that the block-diagonal approximation of the inverse covariance matrix leads to an additive classifier, and demonstrate that accounting for the structure can yield better performance accuracy. Effect of the block size on classification is explored, and a class of asymptotically equivalent structure approximations in a high-dimensional setting is specified. We suggest a variable selection at the block level and investigate properties of this procedure in growing dimension asymptotics. We present a consistency result on the feature selection procedure, establish asymptotic lower an upper bounds for the fraction of separative blocks and specify constraints under which the reliable classification with block-wise feature selection can be performed. The relevance and benefits of the proposed approach are illustrated on both simulated and real data. [ABSTRACT FROM PUBLISHER]

To evaluate the clinical utility of new risk markers, a crucial step is to measure their predictive accuracy with prospective studies. However, it is often infeasible to obtain marker values for all study participants. The nested case-control (NCC) design is a useful cost-effective strategy for such settings. Under the NCC design, markers are only ascertained for cases and a fraction of controls sampled randomly from the risk sets. The outcome dependent sampling generates a complex data structure and therefore a challenge for analysis. Existing methods for analyzing NCC studies focus primarily on association measures. Here, we propose a class of nonparametric estimators for commonly used accuracy measures. We derived asymptotic expansions for accuracy estimators based on both finite population and Bernoulli sampling and established asymptotic equivalence between the two. Simulation results suggest that the proposed procedures perform well in finite samples. The new procedures were illustrated with data from the Framingham Offspring study. [ABSTRACT FROM AUTHOR]

Demographic corrections for cognitive tests should improve classification accuracy by reducing age or education biases, but empirical support has been equivocal. Using a simulation procedure, we show that creating moderate or extreme skewness in cognitive tests compromises the classification accuracy of demographic corrections, findings that appear replicated within clinical data for the few neuropsychological test scores with an extreme degree of skew. For most neuropsychological tests, the dementia classification accuracy of raw and demographically corrected scores was equivalent. These findings suggest that the dementia classification accuracy of demographic corrections is robust to slight degrees of skew (i.e., skewness <1.5). [ABSTRACT FROM AUTHOR]

A criterion-groups validation was used to determine the classification accuracy of the Seashore Rhythm Test (SRT) and Speech Sounds Perception Test (SSPT) in detecting malingered neurocognitive dysfunction (MND) in traumatic brain injury (TBI). TBI patients were classified into the following groups: (1) Mild TBI Not-MND (n = 24); (2) Mild TBI MND (n = 27); and (3) Moderate/Severe TBI Not-MND (n = 23). A sample of 90 general clinical patients was utilized for comparison. Results showed that both SRT correct and SSPT errors differentiated malingerers from non-malingerers in the Mild TBI sample. At 96% specificity, sensitivities were 37% for SRT correct and 59% for SSPT errors. Joint classification accuracy showed that the best accuracy was achieved when using a cut-off associated with a 4% false positive error rate in the Mild TBI sample. Specificity was considerably lower in the Moderate/Severe TBI and General Clinical groups. The clinical application of these findings is discussed. [ABSTRACT FROM AUTHOR]

Reliability is a problem inherent in all educational assessments, but the amount of attention this particular problem should be given is related to the function and use of the assessment. In this article, classification accuracy is put forward as a conceptualization of reliability that is meaningful for a large number of educational assessments. Furthermore, a fairly transparent method of estimating the magnitude of classification accuracy is developed. The article examines how the classification accuracy of a Swedish national test in mathematics is affected by changes in the test that are motivated by other validity concerns. The inquiry is based on the results on 1201 students participating in a Swedish national test in mathematics. The results indicate that there can be a significant trade-off between reliability and other aspects of validity. [ABSTRACT FROM AUTHOR]