A dynamic understanding of customer behavior processes based on clustering and sequence mining

Seret, Alex; Broucke, Seppe K. L. M. vanden; Baesens, Bart; Vanthienen, Jan

doi:10.1016/j.eswa.2014.01.022

Cited by 25 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We believe that embedded feature selection can be very useful in domains such as business analytics (e.g. in customer segmentation tasks), where there are important collection costs linked to each attribute, while the understanding of the underlying behavior of the customers is of primary interest [30]. If no good clustering can be obtained with the candidate variables, a hybrid two-step clustering strategy is suggested, in which features are first filtered out via approaches such as SPEC until a satisfactory partition is found, while KPKM can be used to further polish this clustering.…”

Section: Discussionmentioning

confidence: 99%

Kernel Penalized K-means: A feature selection method based on Kernel K-means

Maldonado

Carrizosa

Weber

2015

Information Sciences

View full text Add to dashboard Cite

Artículo de publicación ISIWe present an unsupervised method that selects the most relevant features using an embedded strategy while maintaining the cluster structure found with the initial feature set. It is based on the idea of simultaneously minimizing the violation of the initial cluster structure and penalizing the use of features via scaling factors. As the base method we use Kernel K-means which works similarly to K-means, one of the most popular clustering algorithms, but it provides more flexibility due to the use of kernel functions for distance calculation, thus allowing the detection of more complex cluster structures. We present an algorithm to solve the respective minimization problem iteratively, and perform experiments with several data sets demonstrating the superior performance of the proposed method compared to alternative approaches.FONDECYT project 11121196 1140831 Complex Engineering Systems Institute ICM: P-05-004-F CONICYT: FB016 Ministerio de Economia y Competitividad MTM2012-36163-C06-03 Junta de Andalucia P11-FQM-7603 FQM 32

show abstract

Section: Discussionmentioning

confidence: 99%

Kernel Penalized K-means: A feature selection method based on Kernel K-means

Maldonado

Carrizosa

Weber

2015

Information Sciences

View full text Add to dashboard Cite

show abstract

“…Qian and Soopramanien [18] showed the emerging markets are to be increasingly important for many companies and it is not surprising to see that an increasing number of new products, especially technology products, are now being launched in these markets fairly quickly after they are launched in Western markets. Seret et al [19] studied that a novel approach enabling the exploratory understanding of the dynamics inherent in the capture of customers' data at different points in time is outlined. Szymczyk and Kamiński [20] discussed the dynamics of innovation diffusion for heterogeneous consumers.…”

Section: Designsmentioning

confidence: 99%

Analysis of Dynamical Marketing Strategy Modeling for Small Modular Reactor (SMR)

Baek¹,

Woo²,

Jang³

2022

IJETAE

View full text Add to dashboard Cite

Using the Hubbert and System Dynamics (SD) methods, the commercialized South Korean Small Modular Reactor (SMR), the System-integrated Modular Advanced Reactor (SMART), is studied for the marking assessment using the non-linear modeling. It is shown for the calculation modeling of the Hubbert method where the period is from 2015 to 2030, because the SMR is expected to be totaled as many as 180 in 2030. However, this period is just an estimated value for this modeling. The uncertainty is expressed in the SD modeling with the described variables. In addition, several variables considered method has a different assessment trend comparing to the Hubbert equation assessment. There are some results as the SMR product, the cumulative product, and the product in maximum time. In the initial stage of the Hubbert modeling, the Product is near 2.0 which could mean the SMART export to Saudi Arabia. Some strategies to nuclear renaissance are discussed including the safety and security technology in nuclear industry with a highest priority. Furthermore, the free-human error technologies have been developed with the robotics technology.

show abstract

“…On the one hand, other data mining techniques as dynamic clustering approaches (see e.g. Seret et al, 2014a;Peters, 2012) or recommender systems can be applied and combined with the output of this segmentation, hence using it as a new knowledge source for future analyses. On the other hand, the current segmentation could also be enriched by considering new data sources while building on the original segmentation as illustrated in the next section.…”

Section: Original Segmentation Of the Customer Basementioning

confidence: 99%

“…By capturing these movements relatively to the xed structure, the dynamics of the customers can thus be explored (see e.g. Seret et al, 2014a). On the other hand, one could consider the clustering structure as an organic component that can evolve and change as a reaction to some triggers and decisions.…”

Section: Introductionmentioning

confidence: 99%

Identifying next relevant variables for segmentation by using feature selection approaches

Seret

Maldonado

Baesens

2015

Expert Systems with Applications

Self Cite

View full text Add to dashboard Cite

Data mining techniques are widely used by researchers and companies in order to solve problems in a myriad of domains. While these techniques are being adopted and used in daily activities, new operational challenges are encountered concerning the steps following this adoption. In this paper, the problem of updating and improving an existing clustering model by adding relevant new variables is studied. A relevant variable is here dened as a feature which is highly correlated with the current structure of the data, since our main goal is to improve the model by adding new information to the current segmentation, but without modifying it signicantly. For this purpose, a general framework is proposed, and subsequently applied in a real business context involving an event organizer facing this problem. Based on extensive experiments based on real data, the performance of the proposed approach is compared to existing methods using dierent evaluation metrics, leading to the conclusion that the proposed technique is performing better for this specic problem.

show abstract

A dynamic understanding of customer behavior processes based on clustering and sequence mining

Cited by 25 publications

References 17 publications

Kernel Penalized K-means: A feature selection method based on Kernel K-means

Kernel Penalized K-means: A feature selection method based on Kernel K-means

Analysis of Dynamical Marketing Strategy Modeling for Small Modular Reactor (SMR)

Identifying next relevant variables for segmentation by using feature selection approaches

Contact Info

Product

Resources

About