A robust and scalable clustering algorithm for mixed type attributes in large database environment

Chiu, Tom; Fang, Dongping; Chen, John; Wang, Yao; Jeris, Christopher

doi:10.1145/502512.502549

Cited by 515 publications

(442 citation statements)

References 6 publications

Supporting

Mentioning

399

Contrasting

Unclassified

Order By: Relevance

“…To classify the different levels of CVCs' strategic and financial investment motivation, we employed cluster analysis to identify mutually exclusive segments of CVCs with a comparable investment motivation (Chiu et al 2001). The clustering method used is based on a two-step procedure, where subclusters are initially defined and subsequently merged until an optimal number of clusters is reached.…”

Section: Clustering Cvcs Based On Their Investment Motivationmentioning

confidence: 99%

A world of difference? The impact of corporate venture capitalists’ investment motivation on startup valuation

et al. 2017

View full text Add to dashboard Cite

Corporate venture capital (CVC) investors are regularly painted with the same brush, a fact underscored by the often observed belief in the extant literature that corporate venture capitalists (CVCs) form a homogeneous group. In contrast to this simplifying perspective, this paper categorizes CVCs into subgroups by examining their levels of strategic and financial investment motivation using computer-aided text analysis and cluster analysis. To validate the resulting clusters, this paper studies the impact of CVC type on startup valuation from an intra-group perspective by applying hierarchical linear modeling, thus illustrating which particular investment motivation might be preferable to others in the context of negotiating valuations. An empirical analysis of 52 CVC mission statements and 147 startup valuations between January 2009 and January 2016 revealed that first, CVCs with a strategic investment motivation assign lower startup valuations than CVCs with an analytic motivation that have moderate levels of the two scrutinized dimensions, suggesting that entrepreneurs trade off these CVCs' value-adding contributions against a valuation discount; second, CVCs with an unfocused investment motivation pay significantly higher purchase prices, thus supporting the hypothesis that they have a so-called liability of vacillation; and third, the valuations of CVCs with a financial investment motive are not significantly different from JEL Classification C12 Á C30 Á C38 Á G24 Á G32 Á M13

show abstract

Section: Clustering Cvcs Based On Their Investment Motivationmentioning

confidence: 99%

A world of difference? The impact of corporate venture capitalists’ investment motivation on startup valuation

et al. 2017

View full text Add to dashboard Cite

show abstract

“…This method is based on a distance measure that enables data with both categorical and continuous attributes to be clustered. This distance measure is derived from a probabilistic model in which the distance between two clusters is equivalent to the decrease in log-likelihood function as a result of merging (Chiu, Fang, Chen, Wang, & Jeris, 2001). …”

Section: Phase 2: Research Questionmentioning

confidence: 99%

“…The BIC is known as one of the most useful and objective selection criteria, because it essentially avoids the arbitrariness in traditional clustering techniques (Norusis, 2003). In addition, both background noise and outliers can be iden-tified and screened out in the algorithm (Chiu et al, 2001).…”

Section: Phase 2: Research Questionmentioning

confidence: 99%

Lessons learned from i-mode: What makes consumers click wireless banner ads?

Okazaki

2007

Computers in Human Behavior

View full text Add to dashboard Cite

This paper attempts to propose a structural model that integrates various factors influencing atti-tude towards wireless banner ads and intention to access them. This model is applied to empirical data of Japanese mobile users sampled in the greater Tokyo area. First, structural equation model-ling is used to test the baseline model. The results show that the model explains mobile users' perceptual antecedents and consequences well, with all structural paths statistically significant. Second, in the attempt to identify different mobile user groups, a probabilistic cluster analysis is performed. This results in three-cluster groups, consisting of (1) housewives and parttimers, (2) mid-dle-aged white-collar workers and professionals, and (3) students and ''parasite singles''. Finally, multigroup analysis is used to examine whether the model operates invariantly across the three-clus-ter groups. The results indicate significant differences in the paths associated with consumer innova-tiveness and perceived entertainment between the groups. In closing, managerial implications and future research directions are discussed, while important limitations are recognised.

show abstract

“…The TURN* [4][2] algorithm locates the knee of a curve by location the point where the 2 nd derivative increases above a user specified threshold. A variant [2] of the BIRCH [23] algorithm uses a mixture of the Bayesian Information Criterion (BIC) and the ratio-change between inter-cluster distance and the number of clusters.…”

mentioning

confidence: 99%

Determining the number of clusters/segments in hierarchical clustering/segmentation algorithms

Salvador

Chan

16th IEEE International Conference on Tools With Artificial Intelligence

444

370

View full text Add to dashboard Cite

Many clustering and segmentation algorithms both suffer from the limitation that the number of clusters/segments are specified by a human user. It is often impractical to expect a human with sufficient domain knowledge to be available to select the number of clusters/segments to return. In this paper, we investigate techniques to determine the number of clusters or segments to return from hierarchical clustering and segmentation algorithms. We propose an efficient algorithm, the L method, that finds the "knee" in a '# of clusters vs. clustering evaluation metric' graph. Using the knee is well-known, but is not a particularly well-understood method to determine the number of clusters. We explore the feasibility of this method, and attempt to determine in which situations it will and will not work. We also compare the L method to existing methods based on the accuracy of the number of clusters that are determined and efficiency. Our results show favorable performance for these criteria compared to the existing methods that were evaluated.

show abstract

A robust and scalable clustering algorithm for mixed type attributes in large database environment

Cited by 515 publications

References 6 publications

A world of difference? The impact of corporate venture capitalists’ investment motivation on startup valuation

A world of difference? The impact of corporate venture capitalists’ investment motivation on startup valuation

Lessons learned from i-mode: What makes consumers click wireless banner ads?

Determining the number of clusters/segments in hierarchical clustering/segmentation algorithms

Contact Info

Product

Resources

About