Comparative Assessment of Hierarchical Clustering Methods for Grouping in Singular Spectrum Analysis

Singular spectrum analysis (SSA) is a popular filtering and forecasting method that is used in a wide range of fields such as time series analysis and signal processing. A commonly used approach to identify the meaningful components of a time series in the grouping step of SSA is the utilization of the visual information of eigentriples. Another supplementary approach is that of employing an algorithm that performs clustering based on the dissimilarity matrix defined by weighted correlation between the components of a time series. The SSA literature search revealed that no investigation has compared the various clustering methods. The aim of this paper was to compare the effectiveness of different hierarchical clustering linkages to identify the appropriate groups in the grouping step of SSA. The comparison was performed based on the corrected Rand (CR) index as a comparison criterion that utilizes various simulated series. It was also demonstrated via two real-world time series how one can proceed, step-by-step, to conduct grouping in SSA using a hierarchical clustering method. This paper is supplemented with accompanying R codes.

Vaginal microbiome topic modeling of laboring Ugandan women with and without fever

The composition of the maternal vaginal microbiome influences the duration of pregnancy, onset of labor, and even neonatal outcomes. Maternal microbiome research in sub-Saharan Africa has focused on non-pregnant and postpartum composition of the vaginal microbiome. Here we aimed to illustrate the relationship between the vaginal microbiome of 99 laboring Ugandan women and intrapartum fever using routine microbiology and 16S ribosomal RNA gene sequencing from two hypervariable regions (V1–V2 and V3–V4). To describe the vaginal microbes associated with vaginal microbial communities, we pursued two approaches: hierarchical clustering methods and a novel Grades of Membership (GoM) modeling approach for vaginal microbiome characterization. Leveraging GoM models, we created a basis composed of a preassigned number of microbial topics whose linear combination optimally represents each patient yielding more comprehensive associations and characterization between maternal clinical features and the microbial communities. Using a random forest model, we showed that by including microbial topic models we improved upon clinical variables to predict maternal fever. Overall, we found a higher prevalence of Granulicatella, Streptococcus, Fusobacterium, Anaerococcus, Sneathia, Clostridium, Gemella, Mobiluncus, and Veillonella genera in febrile mothers, and higher prevalence of Lactobacillus genera (in particular L. crispatus and L. jensenii), Acinobacter, Aerococcus, and Prevotella species in afebrile mothers. By including clinical variables with microbial topics in this model, we observed young maternal age, fever reported earlier in the pregnancy, longer labor duration, and microbial communities with reduced Lactobacillus diversity were associated with intrapartum fever. These results better defined relationships between the presence or absence of intrapartum fever, demographics, peripartum course, and vaginal microbial topics, and expanded our understanding of the impact of the microbiome on maternal and potentially neonatal outcome risk.

Scientific and methodological approaches to modeling the optimal strategy for increasing the competitiveness of pharmacy chains of different sizes

The aim of the work is to develop scientific and methodological approaches to modelling the optimal strategy to increase the competitiveness of pharmacy chains (PC), which belong to different clusters. Materials and methods. The algorithm for determining the optimal strategy for increasing the competitiveness of PC for different clusters using the method of constructing a decision tree and cluster analysis is proposed. To solve this problem, an expert survey of more than 400 pharmacy managers, who were part of the PC of different sizes, was previously conducted. According to the results of an expert survey using hierarchical clustering methods based on the values of 13 input variables - scores of the strengths of the competitiveness of the PC, three clusters of networks were identified, each of which proposed its own algorithm for modelling the optimal strategy of competitiveness. Results. Using modern economic and mathematical tools, the distribution of PC depending on their size into clusters for modelling the dynamics of competitiveness is substantiated. Indicators are identified, which show a significant difference between clusters, which was taken into account in the process of modelling and selection of the optimal strategy to increase the competitiveness of PC. It is established that the biggest negative impact on the strategy of increasing the competitiveness of small networks has a slow response to changes in market conditions, the biggest positive impact – the availability of additional services in the networks; for medium PC the most important factors influencing the level of competitiveness are the location of pharmacies and competent management; for large PC – the use of modern automated management programs, the level of efficiency of the marketing complex and location features. The algorithm of the generalized model of “decision tree” for a choice of optimum strategy of increase of competitiveness depending on the size of PC is constructed. It was found that the following factors are of the greatest importance: the size of the PC, the use of the discount card system, and the least - the speed of response to market changes and the stability of the financial condition. Conclusions. The proposed generalized mathematical model of the “decision tree” allows a reasonable approach to choosing the optimal strategy to increase the competitiveness of PC depending on its size. The assessment of the importance of predictor variables for each cluster of PC allows determining the priority factors in the implementation of measures aimed at implementing the chosen strategy to increase competitiveness

Teleconsultation demand classification and service analysis

Background To enhance teleconsultation management, demands can be classified into different patterns, and the service of each pattern demand can be improved. Methods For the effective teleconsultation classification, a novel ensemble hierarchical clustering method is proposed in this study. In the proposed method, individual clustering results are first obtained by different hierarchical clustering methods, and then ensembled by one-hot encoding, the calculation and division of cosine similarity, and network graph representation. In the built network graph about the high cosine similarity, the connected demand series can be categorized into one pattern. For verification, 43 teleconsultation demand series are used as sample data, and the efficiency and quality of teleconsultation services are respectively analyzed before and after the demand classification. Results The teleconsultation demands are classified into three categories, erratic, lumpy, and slow. Under the fixed strategies, the service analysis after demand classification reveals the deficiencies of teleconsultation services, but analysis before demand classification can’t. Conclusion The proposed ensemble hierarchical clustering method can effectively category teleconsultation demands, and the effective demand categorization can enhance teleconsultation management.

Modelling a Multiplex Brain Network by Local Transfer Entropy

In this work we report on a systematic study of the causal relations in information transfer mechanisms between brain regions under resting condition. The 1000 Functional Connectomes Beijing Zang dataset was used, which includes brain functional images of 180 healthy individuals. We first characterize the information transfer mechanisms by means of Transfer Entropy concepts and, on this basis, propose a set of indexes concerning the whole functional brain network in the frame of a multilayer description. By exploring the influence of a set of states in two given regions at time t (At; Bt.) over the state of one of them at a following time step (Bt+1), a series of time-dependent events can be observed pointing to four kinds of significant interactions, namely:- (de)activation in the same state (ActS); - (de)activation in the oppostive state (ActO);- turn off in the same state (TfS); - turn off in the opposite state (TfO).This leads to four specific rules and to a directional multilayer network based upon four interaction matrices, one for each rule. By hierarchical clustering methods the four rules can be reduced to two sharing some similarities with positive and negative functional connectivity. The global architecture of the four interactions and the features of single nodes were initially explored under stationary conditions. The information transfer mechanisms on the ensuing functional network were studied by specific indexes describing in a multilayer frame the effects of the network structure in several dynamical processes. The healthy subjects database was used to carefully calibrate and validate the proposed approach, whose final aim remains the detection of clinical differences among individuals, as well as among different cognitive states.