Does independent component analysis play a role in unmixing hyperspectral data?

Independent component analysis (ICA) has recently been proposed as a tool to unmix hyperspectral data. ICA is founded on two assumptions: 1) the observed spectrum vector is a linear mixture of the constituent spectra (endmember spectra) weighted by the correspondent abundance fractions (sources); 2)sources are statistically independent. Independent factor analysis (IFA) extends ICA to linear mixtures of independent sources immersed in noise. Concerning hyperspectral data, the first assumption is valid whenever the multiple scattering among the distinct constituent substances (endmembers) is negligible, and the surface is partitioned according to the fractional abundances. The second assumption, however, is violated, since the sum of abundance fractions associated to each pixel is constant due to physical constraints in the data acquisition process. Thus, sources cannot be statistically independent, this compromising the performance of ICA/IFA algorithms in hyperspectral unmixing. This paper studies the impact of hyperspectral source statistical dependence on ICA and IFA performances. We conclude that the accuracy of these methods tends to improve with the increase of the signature variability, of the number of endmembers, and of the signal-to-noise ratio. In any case, there are always endmembers incorrectly unmixed. We arrive to this conclusion by minimizing the mutual information of simulated and real hyperspectral mixtures. The computation of mutual information is based on fitting mixtures of Gaussians to the observed data. A method to sort ICA and IFA estimates in terms of the likelihood of being correctly unmixed is proposed.

Long-term social recognition memory in adult male rats: factor analysis of the social and non-social behaviors

A modified version of the intruder-resident paradigm was used to investigate if social recognition memory lasts at least 24 h. One hundred and forty-six adult male Wistar rats were used. Independent groups of rats were exposed to an intruder for 0.083, 0.5, 2, 24, or 168 h and tested 24 h after the first encounter with the familiar or a different conspecific. Factor analysis was employed to identify associations between behaviors and treatments. Resident rats exhibited a 24-h social recognition memory, as indicated by a 3- to 5-fold decrease in social behaviors in the second encounter with the same conspecific compared to those observed for a different conspecific, when the duration of the first encounter was 2 h or longer. It was possible to distinguish between two different categories of social behaviors and their expression depended on the duration of the first encounter. Sniffing the anogenital area (49.9% of the social behaviors), sniffing the body (17.9%), sniffing the head (3%), and following the conspecific (3.1%), exhibited mostly by resident rats, characterized social investigation and revealed long-term social recognition memory. However, dominance (23.8%) and mild aggression (2.3%), exhibited by both resident and intruders, characterized social agonistic behaviors and were not affected by memory. Differently, sniffing the environment (76.8% of the non-social behaviors) and rearing (14.3%), both exhibited mostly by adult intruder rats, characterized non-social behaviors. Together, these results show that social recognition memory in rats may last at least 24 h after a 2-h or longer exposure to the conspecific.

Unmixing hyperspectral data: independent and dependent component analysis

The development of high spatial resolution airborne and spaceborne sensors has improved the capability of ground-based data collection in the fields of agriculture, geography, geology, mineral identification, detection [2, 3], and classification [4–8]. The signal read by the sensor from a given spatial element of resolution and at a given spectral band is a mixing of components originated by the constituent substances, termed endmembers, located at that element of resolution. This chapter addresses hyperspectral unmixing, which is the decomposition of the pixel spectra into a collection of constituent spectra, or spectral signatures, and their corresponding fractional abundances indicating the proportion of each endmember present in the pixel [9, 10]. Depending on the mixing scales at each pixel, the observed mixture is either linear or nonlinear [11, 12]. The linear mixing model holds when the mixing scale is macroscopic [13]. The nonlinear model holds when the mixing scale is microscopic (i.e., intimate mixtures) [14, 15]. The linear model assumes negligible interaction among distinct endmembers [16, 17]. The nonlinear model assumes that incident solar radiation is scattered by the scene through multiple bounces involving several endmembers [18]. Under the linear mixing model and assuming that the number of endmembers and their spectral signatures are known, hyperspectral unmixing is a linear problem, which can be addressed, for example, under the maximum likelihood setup [19], the constrained least-squares approach [20], the spectral signature matching [21], the spectral angle mapper [22], and the subspace projection methods [20, 23, 24]. Orthogonal subspace projection [23] reduces the data dimensionality, suppresses undesired spectral signatures, and detects the presence of a spectral signature of interest. The basic concept is to project each pixel onto a subspace that is orthogonal to the undesired signatures. As shown in Settle [19], the orthogonal subspace projection technique is equivalent to the maximum likelihood estimator. This projection technique was extended by three unconstrained least-squares approaches [24] (signature space orthogonal projection, oblique subspace projection, target signature space orthogonal projection). Other works using maximum a posteriori probability (MAP) framework [25] and projection pursuit [26, 27] have also been applied to hyperspectral data. In most cases the number of endmembers and their signatures are not known. Independent component analysis (ICA) is an unsupervised source separation process that has been applied with success to blind source separation, to feature extraction, and to unsupervised recognition [28, 29]. ICA consists in finding a linear decomposition of observed data yielding statistically independent components. Given that hyperspectral data are, in given circumstances, linear mixtures, ICA comes to mind as a possible tool to unmix this class of data. In fact, the application of ICA to hyperspectral data has been proposed in reference 30, where endmember signatures are treated as sources and the mixing matrix is composed by the abundance fractions, and in references 9, 25, and 31–38, where sources are the abundance fractions of each endmember. In the first approach, we face two problems: (1) The number of samples are limited to the number of channels and (2) the process of pixel selection, playing the role of mixed sources, is not straightforward. In the second approach, ICA is based on the assumption of mutually independent sources, which is not the case of hyperspectral data, since the sum of the abundance fractions is constant, implying dependence among abundances. This dependence compromises ICA applicability to hyperspectral images. In addition, hyperspectral data are immersed in noise, which degrades the ICA performance. IFA [39] was introduced as a method for recovering independent hidden sources from their observed noisy mixtures. IFA implements two steps. First, source densities and noise covariance are estimated from the observed data by maximum likelihood. Second, sources are reconstructed by an optimal nonlinear estimator. Although IFA is a well-suited technique to unmix independent sources under noisy observations, the dependence among abundance fractions in hyperspectral imagery compromises, as in the ICA case, the IFA performance. Considering the linear mixing model, hyperspectral observations are in a simplex whose vertices correspond to the endmembers. Several approaches [40–43] have exploited this geometric feature of hyperspectral mixtures [42]. Minimum volume transform (MVT) algorithm [43] determines the simplex of minimum volume containing the data. The MVT-type approaches are complex from the computational point of view. Usually, these algorithms first find the convex hull defined by the observed data and then fit a minimum volume simplex to it. Aiming at a lower computational complexity, some algorithms such as the vertex component analysis (VCA) [44], the pixel purity index (PPI) [42], and the N-FINDR [45] still find the minimum volume simplex containing the data cloud, but they assume the presence in the data of at least one pure pixel of each endmember. This is a strong requisite that may not hold in some data sets. In any case, these algorithms find the set of most pure pixels in the data. Hyperspectral sensors collects spatial images over many narrow contiguous bands, yielding large amounts of data. For this reason, very often, the processing of hyperspectral data, included unmixing, is preceded by a dimensionality reduction step to reduce computational complexity and to improve the signal-to-noise ratio (SNR). Principal component analysis (PCA) [46], maximum noise fraction (MNF) [47], and singular value decomposition (SVD) [48] are three well-known projection techniques widely used in remote sensing in general and in unmixing in particular. The newly introduced method [49] exploits the structure of hyperspectral mixtures, namely the fact that spectral vectors are nonnegative. The computational complexity associated with these techniques is an obstacle to real-time implementations. To overcome this problem, band selection [50] and non-statistical [51] algorithms have been introduced. This chapter addresses hyperspectral data source dependence and its impact on ICA and IFA performances. The study consider simulated and real data and is based on mutual information minimization. Hyperspectral observations are described by a generative model. This model takes into account the degradation mechanisms normally found in hyperspectral applications—namely, signature variability [52–54], abundance constraints, topography modulation, and system noise. The computation of mutual information is based on fitting mixtures of Gaussians (MOG) to data. The MOG parameters (number of components, means, covariances, and weights) are inferred using the minimum description length (MDL) based algorithm [55]. We study the behavior of the mutual information as a function of the unmixing matrix. The conclusion is that the unmixing matrix minimizing the mutual information might be very far from the true one. Nevertheless, some abundance fractions might be well separated, mainly in the presence of strong signature variability, a large number of endmembers, and high SNR. We end this chapter by sketching a new methodology to blindly unmix hyperspectral data, where abundance fractions are modeled as a mixture of Dirichlet sources. This model enforces positivity and constant sum sources (full additivity) constraints. The mixing matrix is inferred by an expectation-maximization (EM)-type algorithm. This approach is in the vein of references 39 and 56, replacing independent sources represented by MOG with mixture of Dirichlet sources. Compared with the geometric-based approaches, the advantage of this model is that there is no need to have pure pixels in the observations. The chapter is organized as follows. Section 6.2 presents a spectral radiance model and formulates the spectral unmixing as a linear problem accounting for abundance constraints, signature variability, topography modulation, and system noise. Section 6.3 presents a brief resume of ICA and IFA algorithms. Section 6.4 illustrates the performance of IFA and of some well-known ICA algorithms with experimental data. Section 6.5 studies the ICA and IFA limitations in unmixing hyperspectral data. Section 6.6 presents results of ICA based on real data. Section 6.7 describes the new blind unmixing scheme and some illustrative examples. Section 6.8 concludes with some remarks.

Atrial fibrillation and minimally invasive coronary artery bypass grafting: risk factor analysis.

Atrial fibrillation (AF) is a frequent arrhythmia after conventional coronary artery bypass grafting. With the advent of minimally invasive technique for left internal mammary artery-left anterior descending coronary artery (LIMA-LAD) grafting, we analyzed the incidence and the risk factors of postoperative AF in this patient population. This prospective study involves all patients undergoing isolated LIMA-LAD grafting with minimally invasive technique between January 1994 and June 2000. Twenty-four possible risk factors for postoperative AF were entered into univariate and multivariate logistic regression analyses. Postoperative AF occurred in 21 of the 90 patients (23.3%) analyzed. Double- or triple-vessel disease was present in 12/90 patients (13.3%). On univariate analysis, right coronary artery disease (p <0.01), age (p = 0.01), and diabetes (p = 0.04) were found to be risk factors for AF. On multivariate analysis, right coronary artery disease was identified as the sole significant risk factor (p = 0.02). In this patient population, the incidence of AF after minimally invasive coronary artery bypass is in the range of that reported for conventional coronary artery bypass grafting. Right coronary artery disease was found to be an independent predictor, and this may be related to the fact that in this patient population the diseased right coronary artery was not revascularized at the time of the surgical procedure. For the same reason, this risk factor may find a broader application to noncardiac thoracic surgery.

Confirmatory factor analysis of an integrated model of psycopathology assessed with MMPI-2-RF and the MCMI-III

We presented an integrated hierarchical model of psychopathology that more accurately captures empirical patterns of comorbidity between clinical syndromes and personality disorders.In order to verify the structural validity of the model proposed, this study aimed to analyze the convergence between the Restructured Clinical (RC) scales and Personality scales (PSY-5) of the MMPI-2-RF and the Clinical Syndrome and Personality Disorder scales of the MCMI-III.The MMPI-2-RF and MCMI-III were administered to a clinical sample of 377 outpatients (167 men and 210 women).The structural hypothesiswas assessed by using a Confirmatory Factor Analytic design with four common superordinate factors. An independent-cluster-basis solution was proposed based on maximum likelihood estimation and the application of several fit indices.The fit of the proposed model can be considered as good and more so if we take into account its complexity.

Dual activation of pathways regulated by steroid receptors and peptide growth factors in primary prostate cancer revealed by Factor Analysis of microarray data

Background: We use an approach based on Factor Analysis to analyze datasets generated for transcriptional profiling. The method groups samples into biologically relevant categories, and enables the identification of genes and pathways most significantly associated to each phenotypic group, while allowing for the participation of a given gene in more than one cluster. Genes assigned to each cluster are used for the detection of pathways predominantly activated in that cluster by finding statistically significant associated GO terms. We tested the approach with a published dataset of microarray experiments in yeast. Upon validation with the yeast dataset, we applied the technique to a prostate cancer dataset. Results: Two major pathways are shown to be activated in organ-confined, non-metastatic prostate cancer: those regulated by the androgen receptor and by receptor tyrosine kinases. A number of gene markers (HER3, IQGAP2 and POR1) highlighted by the software and related to the later pathway have been validated experimentally a posteriori on independent samples. Conclusion: Using a new microarray analysis tool followed by a posteriori experimental validation of the results, we have confirmed several putative markers of malignancy associated with peptide growth factor signalling in prostate cancer and revealed others, most notably ERRB3 (HER3). Our study suggest that, in primary prostate cancer, HER3, together or not with HER4, rather than in receptor complexes involving HER2, could play an important role in the biology of these tumors. These results provide new evidence for the role of receptor tyrosine kinases in the establishment and progression of prostate cancer.

Multiple factor analysis of metachronous upper urinary tract transitional cell carcinoma after radical cystectomy

Transitional cell carcinoma (TCC) of the urothelium is often multifocal and subsequent tumors may occur anywhere in the urinary tract after the treatment of a primary carcinoma. Patients initially presenting a bladder cancer are at significant risk of developing metachronous tumors in the upper urinary tract (UUT). We evaluated the prognostic factors of primary invasive bladder cancer that may predict a metachronous UUT TCC after radical cystectomy. The records of 476 patients who underwent radical cystectomy for primary invasive bladder TCC from 1989 to 2001 were reviewed retrospectively. The prognostic factors of UUT TCC were determined by multivariate analysis using the COX proportional hazards regression model. Kaplan-Meier analysis was also used to assess the variable incidence of UUT TCC according to different risk factors. Twenty-two patients (4.6%). developed metachronous UUT TCC. Multiplicity, prostatic urethral involvement by the bladder cancer and the associated carcinoma in situ (CIS) were significant and independent factors affecting the occurrence of metachronous UUT TCC (P = 0.0425, 0.0082, and 0.0006, respectively). These results were supported, to some extent, by analysis of the UUT TCC disease-free rate by the Kaplan-Meier method, whereby patients with prostatic urethral involvement or with associated CIS demonstrated a significantly lower metachronous UUT TCC disease-free rate than patients without prostatic urethral involvement or without associated CIS (log-rank test, P = 0.0116 and 0.0075, respectively). Multiple tumors, prostatic urethral involvement and associated CIS were risk factors for metachronous UUT TCC, a conclusion that may be useful for designing follow-up strategies for primary invasive bladder cancer after radical cystectomy.

Long-term social recognition memory in adult male rats: factor analysis of the social and non-social behaviors

A modified version of the intruder-resident paradigm was used to investigate if social recognition memory lasts at least 24 h. One hundred and forty-six adult male Wistar rats were used. Independent groups of rats were exposed to an intruder for 0.083, 0.5, 2, 24, or 168 h and tested 24 h after the first encounter with the familiar or a different conspecific. Factor analysis was employed to identify associations between behaviors and treatments. Resident rats exhibited a 24-h social recognition memory, as indicated by a 3- to 5-fold decrease in social behaviors in the second encounter with the same conspecific compared to those observed for a different conspecific, when the duration of the first encounter was 2 h or longer. It was possible to distinguish between two different categories of social behaviors and their expression depended on the duration of the first encounter. Sniffing the anogenital area (49.9% of the social behaviors), sniffing the body (17.9%), sniffing the head (3%), and following the conspecific (3.1%), exhibited mostly by resident rats, characterized social investigation and revealed long-term social recognition memory. However, dominance (23.8%) and mild aggression (2.3%), exhibited by both resident and intruders, characterized social agonistic behaviors and were not affected by memory. Differently, sniffing the environment (76.8% of the non-social behaviors) and rearing (14.3%), both exhibited mostly by adult intruder rats, characterized non-social behaviors. Together, these results show that social recognition memory in rats may last at least 24 h after a 2-h or longer exposure to the conspecific.

The underlying return generating factors for REIT returns: an application of independent component analysis

Multi-factor approaches to analysis of real estate returns have, since the pioneering work of Chan, Hendershott and Sanders (1990), emphasised a macro-variables approach in preference to the latent factor approach that formed the original basis of the arbitrage pricing theory. With increasing use of high frequency data and trading strategies and with a growing emphasis on the risks of extreme events, the macro-variable procedure has some deficiencies. This paper explores a third way, with the use of an alternative to the standard principal components approach – independent components analysis (ICA). ICA seeks higher moment independence and maximises in relation to a chosen risk parameter. We apply an ICA based on kurtosis maximisation to weekly US REIT data using a kurtosis maximising algorithm. The results show that ICA is successful in capturing the kurtosis characteristics of REIT returns, offering possibilities for the development of risk management strategies that are sensitive to extreme events and tail distributions.

Confirmatory factor analysis of different versions of the Body Shape Questionnaire applied to Brazilian university students

This study aimed at evaluating the validity, reliability, and factorial invariance of the complete (34-item) and shortened (8-item and 16-item) versions of the Body Shape Questionnaire (BSQ) when applied to Brazilian university students. A total of 739 female students with a mean age of 20.44 (standard deviation = 2.45) years participated. Confirmatory factor analysis was conducted to verify the degree to which the one-factor structure satisfies the proposal for the BSQ's expected structure. Two items of the 34-item version were excluded because they had factor weights (lambda)< 40. All models had adequate convergent validity (average variance extracted =.43-.58; composite reliability=.85-.97) and internal consistency (alpha =.85-.97). The 8-item B version was considered the best shortened BSQ version (Akaike information criterion = 84.07, Bayes information criterion = 157.75, Browne-Cudeck criterion= 84.46), with strong invariance for independent samples (Delta chi(2)lambda(7)= 5.06, Delta chi(2)Cov(8)= 5.11, Delta chi(2)Res(16) = 19.30). (C) 2014 Elsevier Ltd. All rights reserved.

Statistical shape analysis via principal factor analysis

Statistical shape analysis techniques commonly employed in the medical imaging community, such as active shape models or active appearance models, rely on principal component analysis (PCA) to decompose shape variability into a reduced set of interpretable components. In this paper we propose principal factor analysis (PFA) as an alternative and complementary tool to PCA providing a decomposition into modes of variation that can be more easily interpretable, while still being a linear efficient technique that performs dimensionality reduction (as opposed to independent component analysis, ICA). The key difference between PFA and PCA is that PFA models covariance between variables, rather than the total variance in the data. The added value of PFA is illustrated on 2D landmark data of corpora callosa outlines. Then, a study of the 3D shape variability of the human left femur is performed. Finally, we report results on vector-valued 3D deformation fields resulting from non-rigid registration of ventricles in MRI of the brain.

In search of the internal structure of the processes underlying interval timing in the sub-second and the second range: A confirmatory factor analysis approach

One of the earliest accounts of duration perception by Karl von Vierordt implied a common process underlying the timing of intervals in the sub-second and the second range. To date, there are two major explanatory approaches for the timing of brief intervals: the Common Timing Hypothesis and the Distinct Timing Hypothesis. While the common timing hypothesis also proceeds from a unitary timing process, the distinct timing hypothesis suggests two dissociable, independent mechanisms for the timing of intervals in the sub-second and the second range, respectively. In the present paper, we introduce confirmatory factor analysis (CFA) to elucidate the internal structure of interval timing in the sub-second and the second range. Our results indicate that the assumption of two mechanisms underlying the processing of intervals in the second and the sub-second range might be more appropriate than the assumption of a unitary timing mechanism. In contrast to the basic assumption of the distinct timing hypothesis, however, these two timing mechanisms are closely associated with each other and share 77% of common variance. This finding suggests either a strong functional relationship between the two timing mechanisms or a hierarchically organized internal structure. Findings are discussed in the light of existing psychophysical and neurophysiological data.

Elucidating the internal structure of psychophysical timing performance in the sub-second and second range by utilizing confirmatory factor analysis

The most influential theoretical account in time psychophysics assumes the existence of a unitary internal clock based on neural counting. The distinct timing hypothesis, on the other hand, suggests an automatic timing mechanism for processing of durations in the sub-second range and a cognitively controlled timing mechanism for processing of durations in the range of seconds. Although several psychophysical approaches can be applied for identifying the internal structure of interval timing in the second and sub-second range, the existing data provide a puzzling picture of rather inconsistent results. In the present chapter, we introduce confirmatory factor analysis (CFA) to further elucidate the internal structure of interval timing performance in the sub-second and second range. More specifically, we investigated whether CFA would rather support the notion of a unitary timing mechanism or of distinct timing mechanisms underlying interval timing in the sub-second and second range, respectively. The assumption of two distinct timing mechanisms which are completely independent of each other was not supported by our data. The model assuming a unitary timing mechanism underlying interval timing in both the sub-second and second range fitted the empirical data much better. Eventually, we also tested a third model assuming two distinct, but functionally related mechanisms. The correlation between the two latent variables representing the hypothesized timing mechanisms was rather high and comparison of fit indices indicated that the assumption of two associated timing mechanisms described the observed data better than only one latent variable. Models are discussed in the light of the existing psychophysical and neurophysiological data.

A globe-kutatás kultúraváltozóinak vizsgálata faktoranalízis segítségével (Cultural variables of the GLOBE research computed by factor analysis)

A szervezeti kultúraváltozók nem függetlenek egymástól. A tanulmány a GLOBE társadalmi kultúra, társadalmi értékek és a kettő különbségéből képzett differenciaváltozók faktoranalízisével kísérel meg főfaktorokat meghatározni és azokat értelmezni. / === / The cultural variables are not independent. This paper describes and discusses consolidated cultural variables computed by factor analysis from 9 original GLOBE variables of societal paracices, societal values, and differentiation scales.