scholarly journals Estimating the number of clusters via a corrected clustering instability

2020 ◽  
Vol 35 (4) ◽  
pp. 1879-1894
Author(s):  
Jonas M. B. Haslbeck ◽  
Dirk U. Wulff

Abstract We improve instability-based methods for the selection of the number of clusters k in cluster analysis by developing a corrected clustering distance that corrects for the unwanted influence of the distribution of cluster sizes on cluster instability. We show that our corrected instability measure outperforms current instability-based measures across the whole sequence of possible k, overcoming limitations of current insability-based methods for large k. We also compare, for the first time, model-based and model-free approaches to determining cluster-instability and find their performance to be comparable. We make our method available in the R-package .

1990 ◽  
Vol 29 (03) ◽  
pp. 200-204 ◽  
Author(s):  
J. A. Koziol

AbstractA basic problem of cluster analysis is the determination or selection of the number of clusters evinced in any set of data. We address this issue with multinomial data using Akaike’s information criterion and demonstrate its utility in identifying an appropriate number of clusters of tumor types with similar profiles of cell surface antigens.


2019 ◽  
Vol 4 (1) ◽  
pp. 64-67
Author(s):  
Pavel Kim

One of the fundamental tasks of cluster analysis is the partitioning of multidimensional data samples into groups of clusters – objects, which are closed in the sense of some given measure of similarity. In a some of problems, the number of clusters is set a priori, but more often it is required to determine them in the course of solving clustering. With a large number of clusters, especially if the data is “noisy,” the task becomes difficult for analyzing by experts, so it is artificially reduces the number of consideration clusters. The formal means of merging the “neighboring” clusters are considered, creating the basis for parameterizing the number of significant clusters in the “natural” clustering model [1].


Entropy ◽  
2017 ◽  
Vol 19 (9) ◽  
pp. 452 ◽  
Author(s):  
Serkan Akogul ◽  
Murat Erisoglu

2017 ◽  
Vol 13 (2) ◽  
pp. 1-12 ◽  
Author(s):  
Jungmok Ma

One of major obstacles in the application of the k-means clustering algorithm is the selection of the number of clusters k. The multi-attribute utility theory (MAUT)-based k-means clustering algorithm is proposed to tackle the problem by incorporating user preferences. Using MAUT, the decision maker's value structure for the number of clusters and other attributes can be quantitatively modeled, and it can be used as an objective function of the k-means. A target clustering problem for military targeting process is used to demonstrate the MAUT-based k-means and provide a comparative study. The result shows that the existing clustering algorithms do not necessarily reflect user preferences while the MAUT-based k-means provides a systematic framework of preferences modeling in cluster analysis.


2008 ◽  
Vol 2 (1) ◽  
pp. 65-70 ◽  
Author(s):  
M. Cabello ◽  
J. A. G. Orza ◽  
V. Galiano ◽  
G. Ruiz

Abstract. Backtrajectory differences and clustering sensitivity to the meteorological input data are studied. Trajectories arriving in Southeast Spain (Elche), at 3000, 1500 and 500 m for the 7-year period 2000–2006 have been computed employing two widely used meteorological data sets: the NCEP/NCAR Reanalysis and the FNL data sets. Differences between trajectories grow linearly at least up to 48 h, showing faster growing after 72 h. A k-means cluster analysis performed on each set of trajectories shows differences in the identified clusters (main flows), partially because the number of clusters of each clustering solution differs for the trajectories arriving at 3000 and 1500 m. Trajectory membership to the identified flows is in general more sensitive to the input meteorological data than to the initial selection of cluster centroids.


2021 ◽  
Vol 50 (2) ◽  
pp. 187-209 ◽  
Author(s):  
Florian Pein ◽  
Benjamin Eltzner ◽  
Axel Munk

AbstractAnalysis of patchclamp recordings is often a challenging issue. We give practical guidance how such recordings can be analyzed using the model-free multiscale idealization methodology JSMURF, JULES, and HILDE. We provide an operational manual how to use the accompanying software available as an R-package and as a graphical user interface. This includes selection of the right approach and tuning of parameters. We also discuss advantages and disadvantages of model-free approaches in comparison to hidden Markov model approaches and explain how they complement each other.


2019 ◽  
Author(s):  
Adam Morris ◽  
Fiery Andrews Cushman

The alignment of habits with model-free reinforcement learning (MF RL) is a success story for computational models of decision making, and MF RL has been applied to explain phasic dopamine responses, working memory gating, drug addiction, moral intuitions, and more. Yet, the role of MF RL has recently been challenged by an alternate model---model-based selection of chained action sequences---that produces similar behavioral and neural patterns. Here, we present two experiments that dissociate MF RL from this prominent alternative, and present unconfounded empirical support for the role of MF RL in human decision making. Our results also demonstrate that people are simultaneously using model-based selection of action sequences, thus demonstrating two distinct mechanisms of habitual control in a common experimental paradigm. These findings clarify the nature of habits and help solidify MF RL's central position in models of human behavior.


2020 ◽  
Vol 43 ◽  
Author(s):  
Peter Dayan

Abstract Bayesian decision theory provides a simple formal elucidation of some of the ways that representation and representational abstraction are involved with, and exploit, both prediction and its rather distant cousin, predictive coding. Both model-free and model-based methods are involved.


Sign in / Sign up

Export Citation Format

Share Document