Wireless Propagation Multipaths using Spectral Clustering and Three-Constraint Affinity Matrix Spectral Clustering

This study focused on spectral clustering (SC) and three-constraint affinity matrix spectral clustering (3CAM-SC) to determine the number of clusters and the membership of the clusters of the COST 2100 channel model (C2CM) multipath dataset simultaneously. Various multipath clustering approaches solve only the number of clusters without taking into consideration the membership of clusters. The problem of giving only the number of clusters is that there is no assurance that the membership of the multipath clusters is accurate even though the number of clusters is correct. SC and 3CAM-SC aimed to solve this problem by determining the membership of the clusters. The cluster and the cluster count were then computed through the cluster-wise Jaccard index of the membership of the multipaths to their clusters. The multipaths generated by C2CM were transformed using the directional cosine transform (DCT) and the whitening transform (WT). The transformed dataset was clustered using SC and 3CAM-SC. The clustering performance was validated using the Jaccard index by comparing the reference multipath dataset with the calculated multipath clusters. The results show that the effectiveness of SC is similar to the state-of-the-art clustering approaches. However, 3CAM-SC outperforms SC in all channel scenarios. SC can be used in indoor scenarios based on accuracy, while 3CAM-SC is applicable in indoor and semi-urban scenarios. Thus, the clustering approaches can be applied as alternative clustering techniques in the field of channel modeling.

Improving Order-picking Operations with Precedence Constraints through Efficient Storage Location Assignment

This paper is inspired by a manual picking retail company where shape and weight constraints affect the order-picking process. We proposed an alternative clustering similarity index that considers the similarity, the weight and the shape of products. This similarity index was further incorporated in a storage allocation heuristic procedure to set the location of the products. We test the procedure in a retail company that supplies over 191 stores, in Northern Portugal. When comparing the strategy currently used in the company with this procedure, we found out that our approach enabled a reduction of up to 40% on the picking distance; a percentage of improvement that is 32% higher than the one achieved by applying the Jaccard index, a similarity index commonly used in the literature. This allows warehouses to save time and work faster.

Community structure in the World Trade Network based on communicability distances

In this paper, we investigate the mesoscale structure of the World Trade Network. In this framework, a specific role is assumed by short- and long-range interactions, and hence by any suitably defined network-based distance between countries. Therefore, we identify clusters through a new procedure that exploits Estrada communicability distance and the vibrational communicability distance, which turn out to be particularly suitable for catching the inner structure of the economic network. The proposed methodology aims at finding the distance threshold that maximizes a specific quality function defined for general metric spaces. Main advantages regard the computational efficiency of the procedure as well as the possibility to inspect intercluster and intracluster properties of the resulting communities. The numerical analysis highlights peculiar relationships between countries and provides a rich set of information that can hardly be achieved within alternative clustering approaches.

The Kernel Rough K-Means Algorithm

Background: Clustering is one of the most important data mining methods. The k-means (c-means ) and its derivative methods are the hotspot in the field of clustering research in recent years. The clustering method can be divided into two categories according to the uncertainty, which are hard clustering and soft clustering. The Hard C-Means clustering (HCM) belongs to hard clustering while the Fuzzy C-Means clustering (FCM) belongs to soft clustering in the field of k-means clustering research respectively. The linearly separable problem is a big challenge to clustering and classification algorithm and further improvement is required in big data era. Objective: RKM algorithm based on fuzzy roughness is also a hot topic in current research. The rough set theory and the fuzzy theory are powerful tools for depicting uncertainty, which are the same in essence. Therefore, RKM can be kernelized by the mean of KFCM. In this paper, we put forward a Kernel Rough K-Means algorithm (KRKM) for RKM to solve nonlinear problem for RKM. KRKM expanded the ability of processing complex data of RKM and solve the problem of the soft clustering uncertainty. Methods: This paper proposed the process of the Kernel Rough K-Means algorithm (KRKM). Then the clustering accuracy was contrasted by utilizing the data sets from UCI repository. The experiment results shown the KRKM with improved clustering accuracy, comparing with the RKM algorithm. Results: The classification precision of KFCM and KRKM were improved. For the classification precision, KRKM was slightly higher than KFCM, indicating that KRKM was also an attractive alternative clustering algorithm and had good clustering effect when dealing with nonlinear clustering. Conclusion: Through the comparison with the precision of KFCM algorithm, it was found that KRKM had slight advantages in clustering accuracy. KRKM was one of the effective clustering algorithms that can be selected in nonlinear clustering.

The developmental origins of high grade serous ovarian cancer.

e17063 Background: The still persistent uncertainty in the identification of the cell of origin of high grade serous ovarian cancer (HGSOC), with two candidate originating tissues identified in the distal tract of the fallopian tube (FI) and the surface epithelium of the ovary (OSE), has hampered the identification of clinically relevant molecular features for this disease to be targeted for therapy. This resulted in only a negligible improvement in patient’s care since the introduction of carboplatin-based treatments. Methods: To solve this issue, here we show an innovative method based on the identification of a cell of origin-specific DNA methylation print (OriPrint) that allows the reliable stratification of human primary HGSOC in FI and OSE-originated tumors. Results: We show that this approach is robust to alternative clustering methods and can discriminate tumors derived from each of the two origins across different datasets. Also, we translated these findings on a well-characterized retrospective cohort, showing that the cell of origin significantly impacts patient’s prognosis. Finally, through RNAseq, we unveil origin-specific transcriptional networks compatible with the differential impact on patient’s survival, mainly involved in inflammatory response and cell survival, movement and signaling. Conclusions: Our approach proves for the first time in human primary samples that both origins can give rise to HGSOC, paving the way to a finer molecular characterization of this disease and to the development of more effective therapeutic regimens for improved care for patients.

Artificial Neural Network Optimized Approach for Improving Spatial Cluster Quality of Land Value Zone

A geostatistics practical approach is divided data sample into several groups with certain rules. Then, the data groups are used for spatial interpolation. Furthermore, clustering technique is quite commonly used in order to get distance function between sample data. In this study, Self-Organizing Maps (SOM) optimized by using Learning Vector Quantization (LVQ) especially in distance variance have been implemented. The land value zone datasets in Samarinda, East Kalimantan, Indonesia have been used. This study shows that the SOM optimized by LVQ technique have a good distance variance value in the same cluster than SOM technique. In other words, SOM-LVQ can be alternative clustering technique especially centroid position in clusters.

Alternative Clustering

In the last couple of decades, clustering has become a very crucial research problem in the data mining research community. Clustering refers to the partitioning of data objects such as records and documents into groups or clusters of similar characteristics. Clustering is unsupervised learning, because of unsupervised nature there is no unique solution for all problems. Most of the time complex data sets require explanation in multiple clustering sets. All the Traditional clustering approaches generate single clustering. There is more than one pattern in a dataset; each of patterns can be interesting in from different perspectives. Alternative clustering intends to find all unlike groupings of the data set such that each grouping has high quality and distinct from each other. This chapter gives you an overall view of alternative clustering; it's various approaches, related work, comparing with various confusing related terms like subspace, multi-view, and ensemble clustering, applications, issues, and challenges.