Three-Mode Abstracts, Part J

INDEX

•  
• Jackson, M.T.T. (1988a).
  Analysis of tongue positions: Language-specific and cross-linguistic models. Journal of the Acoustical Society of America, 84, 124-143.
  A factor-analytic model, PARAFAC, has been shown to yield highly significant and interpretable results when applied to the analysis of tongue shapes in particular languages. However, the forms of the solutions differ from language to language. A conceptual model that provides an account for cross-language variability in terms of coordinative structures that may vary in a language-specific manner is developed. This model rationalizes both the success of the PARAFAC procedure within languages and the poor cross-language replicability of PARAFAC solutions. We consider methods for quantatively comparing the results of PARAFAC analyses of data from different languages using data from Icelandic and English as a test case. A cross-language factor solution that accounts for 90% of the variance in tongue position measurements is presented and interpreted in terms of the conceptual model.

• Jackson, M.T.T. (1988b).
  Phonetic theory and cross-linguistic variation in vowel articulation. Unpublished doctoral thesis, University of California, Los Angeles.
  This dissertation addresses a concern of long standing in phonetics - the articulatory description of vowel production. Articulatory measurements based on sagittal x-rays, collected from the literature, of Akan (four speakers), Chinese (three), French (four), Icelandic (two), Spanish (five), and Swedish (four) are analyzed using factor analysis. The results are used to develop an explicit speaker-independent model based on articulatory primes which represent potentially independent articulatory gestures. Each articulatory prime generates a family of articulatory configurations.
  Soms hypotheses about the coordination of articulatory primes are investigated using multiple regression- correlation hypothesis-testing methods. There is between-language variation in the articulatory correlates of vowel production, suggesting that articulatory primes are combined in various ways into coordinative structures. This coordinative organization introduces language-specific functional dependencies between various articulatory primes.
  The dissertation concludes with a discussion of how phonetic organization underlies phonological rules affecting vowel quality. It is suggested that articulatory primes form a better framework for the description of vowel umlaut rules than height and backness, or constriction location and degree.

• Jaffrennou, P.A. (1978).
  Sur l'analyse des familles finies de variables vectorielles. Bases algèbriques et application à la description statistique [On the analysis of finite families of vector variables: Algebraic bases and applications in statistical description]. (Research report). Saint-Etienne, France: University of Saint-Etienne, Department of Mathematics.
  
         
• Jedidi, K., & DeSarbo, W.S. (1991).
  A stochastic multidimensional scaling procedure for the spatial representation of three-mode, three-way pick any/J data. Psychometrika, 56, 471-494.
  This paper presents a new stochastic multidimensional scaling procedure for the analysis of three-mode, three-way pick any/J data. The method provides either a vector or ideal-point model to represent the structure in such data, as well as "floating" model specifications (e.g., different vectors or ideal points for different choice settings), and various reparameterization options that allow the coordinates of ideal points, vectors, or stimuli to be functions of specified background variables. A maximum likelihood procedure is utilized to estimate a joint space of row and column objects, as well as a set of weights depicting the third mode of the data. An algorithm using a conjugate gradient method with automatic restarts is developed to estimate the parameters of the models. A series of Monte Carlo analyses are carried out to investigate the performance of the algorithm under diverse data and model specification conditions, examine the statistical properties of the associated test statistic, and test the robustness of the procedure to departures from the independence assumptions. Finally, a consumer psychology application assessing the impact of situational influences on consumers' choice behavior is discussed.

• Jennrich, R. (1972).
  A generalization of the multidimensional scaling model of Carroll and Chang. UCLA Working Papers in Phonetics, 22, 45-47.
  Proposal to relax assumptions of INDSCAL to allow for individual positioning of the common space axes (= IDIOSCAL). Very brief sections on estimation and computing.

         

• Jiang, J.H., Wu, H.L., Chen, Z.P. & Yu, R.Q. (1999).
  Coupled vectors resolution method for chemometric calibration with three-way data. Analytical Chemistry, 71, 4254-4262.
  A new second-order calibration procedure, the coupled vectors resolution (COVER) method, has been developed. The objective of the method is to seek a couple of vectors that minimize a least-squares criterion. With the knowledge indispensable for quantitation, the method yields direct solutions to various cases of second-order calibration. Moreover, it allows a statistically plausible way to make use of multisample information. In the case of multiple calibration samples, the method uses the calibration samples to resolve the profiles of the analytes in each order, and then calculates the concentrations of the analytes. This offers the advantage that unknown mixtures newly collected can be predicted in a direct manner. In the case of one calibration sample, the method provides an effective way of utilizing the information of spectral profiles of the analytes. Results of simulated experiments and a real analytical example show that the proposed method produces acceptable performance in profile resolution and concentration estimation.

• Jiang, J.H., Wu, H.L., Li, Y. & Yu, R.Q. (1999).
  Alternating coupled vectors resolution (ACOVER) method for trilinear analysis of three-way data. Journal of Chemometrics, 13, 557-578.
  A new method, alternating coupled vectors resolution (ACOVER), is developed for trilinear analysis of a three-way data array. First, based on the least-squares principle, four coupled vectors resolution (COVER) errors are proposed. Second, a procedure which resolves the profiles of each component successively is developed. This characteristic of successive resolution provides a natural way to avoid the two-factor degeneracy. Experimental results show that the ACOVER method has the advantage that the resolved profiles of analytical interest are very stable with respect to the estimated component number when the number of components is chosen to be equal to or greater than the actual model dimensionality. This circumvents the dilemma of determining a proper component number for the model, which is difficult to handle for the PARAFAC algorithm. Moreover, the method has much higher convergence rate than the PARAFAC algorithm.

• Jiang, J. H., Wu, H. L., Li, Y. & Yu, R. Q. (2000).
  Three-way data resolution by alternating slice-wise diagonalization (ASD) method. Journal of Chemometrics, 14, 15-36.
  A new approach, the alternating slice-wise diagonalization (ASD) method, is developed for three-way data resolution. First, based on the least squares principle and the constraints inherent in the resolution of the trilinear model, a criterion, the slice-wise diagonalization (SD) loss, is proposed for trilinear analysis of three-way data. This criterion provides a natural way to avoid the two-factor degeneracy, which is difficult to handle for the PARAFAC algorithm. Second, by alternatingly minimizing the SD loss, a procedure is developed for identifying the parameters of the trilinear model. Experimental results show that the resolved profiles of chemical meaning are very stable with respect to the component number provided that the number is chosen to be equal to or greater than the actual one. This enables the ASD method to achieve resolution without concern about the actual component number. This approach is different from the traditional ones, since the determination of the actual component number is a critical step for conventional chemometric resolution techniques. Moreover, the convergence rate of the algorithm for the ASD method is much higher than that of the PARAFAC algorithm.

• Jiang, T., & Sidiropoulos, N. D. (2004a).
  Kruskal's permutation lemma and the identification of CANDECOMP/PARAFAC and bilinear models with constant modulus constraints. IEEE Transactions on Signal Processing, 52, 2625-2636.
  CANDECOMP/PARAFAC (CP) analysis is an extension of low-rank matrix decomposition to higher-way arrays, which are also referred to as tensors. CP extends and unifies several array signal processing tools and has found applications ranging from multidimensional harmonic retrieval and angle-car-rier estimation to blind multiuser detection. The uniqueness of CP decomposition is not fully understood yet, despite its theoretical and practical significance. Toward this end, we first revisit Kruskal’s Permutation Lemma, which is a cornerstone result in the area, using an accessible basic linear algebra and induction approach. The new proof highlights the nature and limits of the identification process. We then derive two equivalent necessary and sufficient uniqueness conditions for the case where one of the component matrices involved in the decomposition is full column rank. These new conditions explain a curious example provided recently in a previous paper by Sidiropoulos, who showed that Kruskal’s condition is in general sufficient but not necessary for uniqueness and that uniqueness depends on the particular joint pattern of zeros in the (possibly pretransformed) component matrices. As another interesting application of the Permutation Lemma, we derive a similar necessary and sufficient condition for unique bilinear factorization under constant modulus (CM) constraints, thus providing an interesting link to (and unification with) CP.

• Jiang, T., & Sidiropoulos, N. D. (2004b).
  Blind identification of out-of-cell users in DS-CDMA. Eurasip Journal on Applied Signal Processing, 2004, 1212-1224.
  In the context of multiuser detection for the DS-CDMA uplink, out-of-cell interference is usually treated as Gaussian noise, possibly mitigated by overlaying a long random cell code on top of symbol spreading. Different cells use statistically independent long codes, thereby providing means for statistical out-of-cell interference suppression. When the total number of (in-cell plus out-of-cell) users is less than the spreading gain, subspace identification techniques are applicable. If the base station is equipped with multiple antennas, then completely blind identification is possible via three-dimensional low-rank decomposition. This works with more users than spreading and antennas, but a purely algebraic solution is missing. In this paper, we develop an algebraic solution under the premise that the codes of the in-cell users are known. The codes of out-of-cell users and all array steering vectors are unknown. In this pragmatic scenario, we show that in addition to algebraic solution, better identifiability is possible. Our approach yields the best known identifiability result for three-dimensional low-rank decomposition when one of the three component matrices is partially known, albeit noninvertible. Simulations show that the proposed identification algorithm remains close to the pertinent asymptotic (symbol-independent) Cramer-Rao bound, which is also derived herein.

• Jiji, R.D., Andersson, G.G. & Booksh, K.S. (2000).
  Application of PARAFAC for calibration with excitation-emission matrix fluorescence spectra of three classes of environmental pollutants. Journal of Chemometrics, 14, 171-185.
  Parallel factor analysis (PARAFAC) is applied to three calibrations of a field-portable, cuvette-based, single-measurement, excitation-emission matrix fluorometer. In the first example the fluorometer is calibrated based on interactions between a non-fluorescent DDT-type pesticide and a fluorescent dye. Parafac is employed to deconvolve the fluorescence profiles of dissociated and complexed dye states. Calibration is performed based on the intensity of dye-pesticide fluorescence. In the second example, weighted parafac (W-PARAFAC) is applied to determination of three polynuclear aromatic hydrocarbons (PAHs). The weighted algorithm is required to incorporate saturated channels of the CCD detector into the calibration model. In the third example, W-PARAFAC is applied to calibration of two carbamate pesticides. The weighted algorithm is required to account for Rayleigh and Raman scattering overlapping with the fluorescence spectra. For theses three applications, parts-per-trillion to parts-per-billion detection limits are observed in aqueous solutions.

• Jiji, R.D. & Booksh, K.S. (2000).
  Mitigation of Rayleigh and Raman spectral interferences in multiway calibration of excitation-emission matrix fluorescence spectra. Analytical Chemistry, 72, 718-725.
  A weighted parallel factor analysis (W-PARAFAC) model is applied to excitation-emission matrix (EEM) fluorescence spectra of carbamate pesticides to aid with calibration in the presence of Raman scattering. Traditional PARAFAC inefficiently models the Raman scattering, resulting in prediction and calibration errors when a significant background is present. Four different weighting strategies were investigated and compared with subtraction of the appropriate sample background. Using a binary weighting strategy produced superior results, compared with a continuous distribution of weights. Further choice of weighting strategies, which are optimized to include either maximum analyte signal or to exclude a maximum amount of background scattering, is dependent on the degree of overlap and relative signal intensity.

• Jiji, R.D., Cooper, G.A. & Booksh, K.S. (1999).
  Excitation-emission matrix fluorescence based determination of carbamate pesticides and polycyclic aromatic hydrocarbons. Analytica Chimica Acta, 397, 61-72.
  The ability to accurately quantitate trace pesticides and polycyclic aromatic hydrocarbons (PAHs) by excitation-emission matrix (EEM) fluorescence spectroscopy, when coupled with parallel factor analysis (PARAFAC) deconvolution of the EEM spectra, is demonstrated and discussed. Two EEM fluorometers were investigated. One fluorometer, using a cuvette cell sample holder, realized limits of detection of 1.1, 6.6, and 13 ppb for 1-naphthol, carbaryl, and carbofuran in methanol mixtures, respectively. 0.2 ppb limits of detection were also observed for three PAHs. With this fluorometer, the PARAFAC model was employed to resolve the analyte spectra from overlapping fluorescence signals and Raman scattering profiles. Employing a second fluorometer with a fiber-optic probe for remote sampling yielded 10-30 ppb limits of detection for 9 PAHs. The PARAFAC model was required here to resolve the PAH spectra from the instrumental background.

             

• Johnson, K., Ladefoged, P., & Lindau, M. (1993).
  Individual-differences in vowel production. Journal of the Acoustical Society of America, 94, 701- 714.**
  It is often assumed that a relatively small set of articulatory features are universally used in language sound systems. This paper presents a study which tests this assumption. The data are x-ray microbeam pellet trajectories during the production of the vowels of american english by five speakers. Speakers were consistent with themselves from one production of a word to the next, but the articulatory patterns manifested by this homogeneous group were speaker specific. Striking individual differences were found in speaking rate, the production of the tense/lax distinction of english, and in global patterns of articulation. In terms of a task-dynamic model of speech production, these differences suggested that the speakers used different gestural target and stiffness values, and employed different patterns of interarticulator coordination to produce the vowels of american english. The data thus suggest that, at some level of speech motor control, speech production tasks are specified in terms of acoustic output rather than spatiotemporal targets or gestures.

  ---

• Johnson, K., De Juan, A., & Rutan, S.C. (1999).
  Three-way data analysis of pollutant degradation profiles monitored using liquid chromatography-diode array detection. Journal of Chemometrics, 13, 331-341.
  There is a wide variety of environmental contaminants that needs to be better characterized. It is important to know the degradation pathways for these pollutants in order to better understand their effect on the environment. The hydrolysis of Glean^® (chlorsulfuron), a sulfonylurea herbicide, is currently under study in this laboratory. Liquid chromatography coupled with diode array detection (LC-DAD) offers a means of studying these types of reactions. With this approach, three-way data are obtained - absorbance measurements as a function of chromatographic retention time, UV wavelength and reaction time. The direct trilinear decomposition (DTD) algorithm was applied to these data, presuming an internal trilinear structure in the data set. Upon inspection of the results, however, deviations from trilinearity were found that caused chemically meaningless results to be obtained using this approach. This difficulty was addressed by using the three-way multivariate curve resolution-alternating least squares approach (MCR-ALS). This technique allows for the optional use of the trilinearity constraint, and some additional constraints related to the features of the chromatographic and spectral profiles can be included. In the present work the application of this approach has been evaluated for resolving the overlapped responses that arise when measuring herbicide degradation reaction profiles.

• Johnson, K. J., Rose-Pehrsson, S. L., & Morris, R. E. (2004).
  Monitoring diesel fuel degradation by gas chromatography-mass Spectroscopy and chemometric analysis. Energy & Fuels, 18, 844-850.
  Recent advances in the field of chemometrics have provided us with an opportunity to determine if it is possible to extend numerical multivariate pattern recognition techniques beyond simple fuel characterization, to the identification of important compositional features that are related to fuel quality. This can include unique combinations of normally benign constituents that exert an impact on fuel stability. Fuels are ideal candidates for chemometric analysis, because we are often concerned with minute features within a complex compositional matrix. Two potential benefits of this approach are the development of diagnostic and predictive models that can relate fuel composition to quality. We have begun our investigation with studies of gas chromatography-mass spectrometry (GC-MS) data from fuels that have undergone various levels of thermally induced autoxidation. An analysis of variance (ANOVA) feature selection technique has been applied to locate features in the GC-MS data that change from sample to sample, thus allowing for a quick evaluation of how fuel composition is altered during stress. In this manner, evaporative losses, rather than fuel degradation, have been observed to dominate the chemical variations that are produced in naval distillate fuels (NATO F-76) during oven stress at 60 degreesC. Thermal stress in a closed low-pressure reactor (LPR) eliminates evaporative losses, and the chemical changes have been readily observed and modeled. A progressive change in composition during both oven and LPR stress is revealed from multi-way principal component analysis. Decomposition of windowed regions of the GC-MS data via parallel factor analysis provides a means of extracting the mass spectra of individual fuel constituents that change during stress. This illustrates the potential diagnostic capability of multi-way chemometric analysis of GC-MS data.

• Jones, L.E., & Iacobucci, D. (1989).
  The structure of affect and trait judgments of political figures. Multivariate Behavioral Research, 24, 457- 476.
  The roles of affective and cognitive processes in judgment have been the focus of much recent research and theoretical debate. This study was designed to investigate the structure of voters' affective reactions and trait attributions to national political figuers. Three-mode factor analysis was used to determine the structure of the affect and trait scales. Politician factors and subject types were also derived, as were the interrelationships among these modes. Positive and negative affect factors, and affect and trait factors, were distinct but correlated; Democratic and Republican politician factors were uncorrelated. Information on the subject types moderated these relationships.

• Jouan-Rimbaud, D., Massart, D. L., Saby, C. A., & Puel, C. (1998).
  Determination of the representativity between two multidimensional data sets by a comparison of their structure. Chemometrics & Intelligent Laboratory Systems, 40, 129-144.
  An approach for the investigation and comparison of the data structure in the multidimensional space is proposed. It is based on three properties, namely, the direction of the data sets, the variance–covariance of the data points, and the location of the data sets’ centroids. A number of tests have been studied and are presented. It is shown that the combined use of these parameters allows a satisfactory estimation of the representativity between two data sets. Simulated data, as well as real case studies are presented.

• Jørgensen, R. N., Hansen, P. M., & Bro, R. (2006).
  Exploratory study of winter wheat reflectance during vegetative growth using three-mode component analysis. International Journal of Remote Sensing, 27, 919-937.
  Remote sensing is a potentially important source of data for site-specific crop management, providing both spatial and temporal information. Remote?sensed data have been used in nitrogen application strategies to achieve optimal yield potential or grain quality. Our objective was to investigate and describe the use of repeated multispectral measurements and three?mode component analysis in order to interpret measurements on winter wheat grown at different plant densities and nitrogen application strategies. A multi-band spectroradiometer with eight different wavelengths was used three times during vegetative growth. The Tucker3 model was used for the three-mode component analysis (3MPCA). A comparative study of single reflectance wavelengths and selected calculated vegetation indices was performed. Results showed that the 3MPCA technique gave better separation of samples from different treatments than a single principal components analysis (PCA) using a single measurement day. Furthermore, vegetation indices seemed to perform better than single wavebands. The 3MPCA model is capable of extracting important structural information about measurement conditions and cropping history. The temporal repeated measurements provided additional information about average growth rate and possibly also irradiation conditions at the time of measurement.

                       

|Top | Algemene en Gezinspedagogiek - Datatheorie | Centre for Child and Family Studies | Department of Educational Sciences | The Three-Mode Company | Three-Mode bibliography |