Author

# S. Winsberg

Bio: S. Winsberg is an academic researcher from IRCAM. The author has contributed to research in topics: Multidimensional scaling & Interval (mathematics). The author has an hindex of 3, co-authored 5 publications receiving 62 citations.

##### Papers

More filters

••

TL;DR: A new algorithm called I-Scal, based on iterative majorization, that has the advantage that each iteration is guaranteed to improve the solution until no improvement is possible is developed.

46 citations

•

30 Mar 2005

TL;DR: This paper provides a new algorithm called SymScal that is based on iterative majorization that is guaranteed to improve the solution until no improvement is possible and discusses the use of Sym scal on empirical dissimilarity intervals of sounds.

Abstract: Multidimensional scaling aims at reconstructing dissimilarities between pairs of objects by distances in a low dimensional space. However, in some cases the dissimilarity itself is unknown, but the range of the dissimilarity is given. Such fuzzy data fall in the wider class of symbolic data (Bock & Diday, 2000). Denœux and Masson (2002) have proposed to model an interval dissimilarity by a range of the distance defined as the minimum and maximum distance between two rectangles representing the objects. In this paper, we provide a new algorithm called SymScal that is based on iterative majorization. The advantage is that each iteration is guaranteed to improve the solution until no improvement is possible. In a simulation study, we investigate the quality of this algorithm. We discuss the use of SymScal on empirical dissimilarity intervals of sounds.

10 citations

••

01 Jan 2006TL;DR: This paper provides a new algorithm called Hist-Scal using iterative majorization, that is based on an algorithm, I- Scal developed for the case where the dissimilarities are given by a range of values ie an interval (see Groenen et al. (in press).

Abstract: Multidimensional scaling aims at reconstructing dissimilarities between pairs of objects by distances in a low dimensional space. However, in some cases the dissimilarity itself is unknown, but the range, or a histogram of the dissimilarities is given. This type of data fall in the wider class of symbolic data (see Bock and Diday (2000)). We model a histogram of dissimilarities by a histogram of the distances defined as the minimum and maximum distance between two sets of embedded rectangles representing the objects. In this paper, we provide a new algorithm called Hist-Scal using iterative majorization, that is based on an algorithm, I-Scal developed for the case where the dissimilarities are given by a range of values ie an interval (see Groenen et al. (in press)). The advantage of iterative majorization is that each iteration is guaranteed to improve the solution until no improvement is possible. We present the results on an empirical data set on synthetic musical tones.

8 citations

••

01 Jan 2007TL;DR: In this paper, a new algorithm called 3WaySym-Scal using iterative majorization is proposed, which is based on an algorithm, I-scal developed for the two-way case where the dissimilarities are given by a range of values ie an interval.

Abstract: Multidimensional scaling aims at reconstructing dissimilarities between pairs of objects by distances in a low dimensional space. However, in some cases the dissimilarity itself is not known, but the range, or a histogram of the dissimilarities is given. This type of data fall in the wider class of symbolic data (see Bock and Diday (2000)). We model three-way two-mode data consisting of an interval of dissimilarities for each object pair from each of K sources by a set of intervals of the distances defined as the minimum and maximum distance between two sets of embedded rectangles representing the objects. In this paper, we provide a new algorithm called 3WaySym-Scal using iterative majorization, that is based on an algorithm, I-Scal developed for the two-way case where the dissimilarities are given by a range of values ie an interval (see Groenen et al. (2006)). The advantage of iterative majorization is that each iteration is guaranteed to improve the solution until no improvement is possible. We present the results on an empirical data set on synthetic musical tones.

•

TL;DR: This paper model three-way two-mode data consisting of an interval of dissimilarities for each object pair from each of K sources by a set of intervals of the distances defined as the minimum and maximum distance between two sets of embedded rectangles representing the objects.

Abstract: Multidimensional scaling aims at reconstructing dissimilarities between pairs of objects by distances in a low dimensional space. However, in some cases the dissimilarity itself is not known, but the range, or a histogram of the dissimilarities is given. This type of data fall in the wider class of symbolic data (see Bock and Diday (2000)). We model three-way two-mode data consisting of an interval of dissimilarities for each object pair from each of K sources by a set of intervals of the distances defined as the minimum and maximum distance between two sets of embedded rectangles representing the objects. In this paper, we provide a new algorithm called 3WaySym-Scal using iterative majorization, that is based on an algorithm, I-Scal developed for the two-way case where the dissimilarities are given by a range of values ie an interval (see Groenen et al. (2006)). The advantage of iterative majorization is that each iteration is guaranteed to improve the solution until no improvement is possible. We present the results on an empirical data set on synthetic musical tones.

##### Cited by

More filters

••

TL;DR: This new method shows the importance of range information in prediction performance as well as the use of inequality constraints to ensure mathematical coherence between the predicted values of the lower and upper boundaries of the interval value of the dependent variable.

188 citations

••

183 citations

••

TL;DR: In spite of a growing literature concerning the development and application of fuzzy techniques in statistical analysis, the need is felt for a more systematic insight into the potentialities of cross fertilization between Statistics and Fuzzy Logic.

129 citations

••

TL;DR: This paper introduces three approaches to forecasting interval-valued time series based on multilayer perceptron (MLP) neural networks and Holt’s exponential smoothing methods, respectively.

117 citations

••

TL;DR: In this article, a multidimensional scaling technique devised by Carroll and Chang [Psychometrika, 1970] offers a solution to the rotational problem arrived at by capitalizing on individual differences: the axes derived are those that best account for variations among individuals in terms of a differential weighting model.

Abstract: An experimenter who performs a multidimensional analysis of speech perception data must deal with two problems: (1) differences among subjects in the data and (2) rotation of axes in the solution. The rotational problem is crucial when one seeks to determine whether the “true” perceptual dimensions correspond to a particular set of acoustic measures or articulatory features of the stimuli. A new multidimensional scaling technique devised by Carroll and Chang [Psychometrika, (1970) (to be published)] offers a solution to the rotational problem arrived at by capitalizing on individual differences: the axes derived are those that best account for variations among individuals in terms of a differential weighting model. This paper illustrates the promise this method holds for speech perception studies by an example: data on dissimilarities among nine Swedish vowels published by Hanson [Ericsson Technics 23, 3–175 (1967)] are re‐analyzed. Whereas the stimulus set represented three distinctive features, the published solution recovered only two. The Carroll‐Chang solution is interpretable in terms of all three features and formant frequencies as well. Additional procedures show that the two solutions differ chiefly in rotation and illustrate how other (superficially rational) rotational criteria can lead to less informative solutions.

74 citations