Search results

This study compares the performance of two types of clustering methods, time-based and non-time-based clustering, in the identification of river discharge patterns at the Johor River basin during the northeast monsoon season. Time-based clustering is represented by employing dynamic time warping (DTW) dissimilarity measure, whereas non-time-based clustering is represented by employing Euclidean dissimilarity measure in analysing the Johor River discharge data. In addition, we combine each of these clustering methods with a frequency domain representation of the discharge data using Discrete Fourier Transform (DFT) to see if such transformation affects the clustering results. The clustering quality from the hierarchical data structures of the identified river discharge patterns for each of the methods is measured by the Cophenetic Correlation Coefficient (CPCC). The results from the time-based clustering using DTW based on DFT transformation show a higher CPCC value as compared to that of non-time-based clustering methods.

Given the recent explosion of interest in streaming data and online algorithms, clustering of time series subsequences has received much attention. In this work we make a surprising claim. Clustering of time series subsequences is completely meaningless. More concretely, clusters extracted from these time series are forced to obey a certain constraint that is pathologically unlikely to be satisfied by any dataset, and because of this, the clusters extracted by any clustering algorithm are essentially random. While this constraint can be intuitively demonstrated with a simple illustration and is simple to prove, it has never appeared in the literature. We can justify calling our claim surprising, since it invalidates the contribution of dozens of previously published papers. We will justify our claim with a theorem, illustrative examples, and a comprehensive set of experiments on reimplementations of previous work.

This chapter acknowledges the difficulties in the empirical study of social ventures – broadly defined as market-driven ventures that produce social change – that arise from the vast differences among social venture firms in terms of missions, goals, identities, strategies, and structures. In an effort to improve research in this area and advance the field of social entrepreneurship, the authors advocate approaching social ventures from a configurational perspective.

This chapter begins with a discussion of what social ventures are and why organizational configurations – sets of firms that are similar across key characteristics – may be an appropriate perspective to utilize. Then, two methods – cluster analysis and set-theoretic analysis – are discussed in detail as ways to approach the study of social venture configurations. Details include descriptions of the techniques, instructions for use, examples, and limitations for each.

This chapter identifies research opportunities using configurations approaches in social venture research. Substantial possibilities for multilevel and temporally based research are discussed in depth.

A configurational approach can address the incongruence and non-findings in current social venture research and offers new opportunities for future research.

This paper attempts to evaluate whether the set of NAFTA countries (the U.S., Canada and Mexico) should adopt the same currency. The theoretical basis for the paper is the optimal currency area theory which suggests that countries or regions that experience similar business cycles can gain advantages in adopting the same currency. The statistical methodology used in the paper to evaluate whether states or provinces have similar business cycle correlations is model-based cluster analysis, a recently-developed method to group data in the applied statistics literature.

Multiple linear regression (MLR) is a commonly used statistical technique to predict future values. In this paper, we examine the situation in which a given time series dataset contains numerous observations of important predictor variables that can effectively be classified into groups based on their values. In such situations, cluster analysis is often employed to improve the MLR models predictive accuracy, usually by creating separate regressions for each cluster. We introduce a novel approach in which we use the clusters and cluster centroids as input data for the predictor variables to improve the predictive accuracy of the MLR model. We illustrate and test this approach with a real dataset on fleet maintenance.

Constructing and evaluating behavioral science models is a complex process. Decisions must be made about which variables to include, which variables are related to each other, the functional forms of the relationships, and so on. The last 10 years have seen a substantial extension of the range of statistical tools available for use in the construction process. The progress in tool development has been accompanied by the publication of handbooks that introduce the methods in general terms (Arminger et al., 1995; Tinsley & Brown, 2000a). Each chapter in these handbooks cites a wide range of books and articles on specific analysis topics.

We provide an overview of how artificial intelligence is transforming the identification, structuring, and prioritization of customer needs – known as the voice of the customer (VOC). First, we summarize how the VOC helps firms gain insights on using user-generated data. Second, we discuss the types of user-generated data and the challenges associated with analyzing each type of data. Third, we describe common methods, matched to the firms' goals and the structure of the data, that are used to analyze the VOC. Fourth, and most importantly, we map the methods to relevant applications, providing guidance to select the appropriate method to address the desired research questions.