Clustering of data has numerous applications and has been studied extensively. Though most of the algorithms in the literature are sequential, many parallel algorithms have also been designed. In this paper, we present parallel algorithms with better performance than known algorithms. We consider algorithms that work well in the worst case as well as algorithms with good expected performance.

Efficient Parallel Hierarchical Clustering Algorithms.

Single-cell analysis through mass cytometry has become an increasingly important tool for immunologists to study the immune system in health and disease. Mass cytometry creates a high-dimensional description vector for single cells by time-of-flight measurement. Recently, t-Distributed Stochastic Neighborhood Embedding (t-SNE) has emerged as one of the state-of-the-art techniques for the visualization and exploration of single-cell data. Ever increasing amounts of data lead to the adoption of Hierarchical Stochastic Neighborhood Embedding (HSNE), enabling the hierarchical representation of the data. Here, the hierarchy is explored selectively by the analyst, who can request more and more detail in areas of interest. Such hierarchies are usually explored by visualizing disconnected plots of selections in different levels of the hierarchy. This poses problems for navigation, by imposing a high cognitive load on the analyst. In this work, we present an interactive summary-visualization to tackle this problem. CyteGuide guides the analyst through the exploration of hierarchically represented single-cell data, and provides a complete overview of the current state of the analysis. We conducted a two-phase user study with domain experts that use HSNE for data exploration. We first studied their problems with their current workflow using HSNE and the requirements to ease this workflow in a field study. These requirements have been the basis for our visual design. In the second phase, we verified our proposed solution in a user evaluation.

/pdf/cyteguide-visual-guidance-for-hierarchical-single-cell-20em4d8be2.pdf

CyteGuide: Visual Guidance for Hierarchical Single-Cell Analysis

Small multiples can visually enforce comparisons of changes or differences among objects, revealing potential patterns by providing different views. According to the analyzing requirements in food safety fields and characteristics of pesticide residue detection data, in this paper, we propose a novel visualization approach to explore and analyze the time-varying hierarchical data, which is called ordered small multiple treemaps (OSMT). Inspired by the thought of querying an array by rows or columns, OSMT makes it possible to locate a specific node in the treemap layout by using a unique location 2-tuple and keep a relative stable order of nodes in the layout while we detecting temporal patterns. This algorithm enables the visual representation of the node values varying with time, preserving the hierarchical relationships among nodes in the meanwhile. Based on some interaction techniques (filtering, selecting, highlighting and zooming, etc.), OSMT can help users find some specific changes more easily and thus make corresponding decisions with more efficiency. Besides, we also propose a new metric called TVA (Ability of tracking time-varying data in treemap) with a purpose of evaluating different kinds of treemap layout algorithms from the aspect of the difficulty level for tracking time-varying nodes in the overall layout. Finally, our technique's applicability is demonstrated on the pesticide residues detection results dataset in this study.

Ordered small multiple treemaps for visualizing time-varying hierarchical pesticide residue data

Maximum residue limit (MRL) standard which specifies the highest level of every pesticide residue in different agricultural products plays a critical role in food safety. However, such standards which related to the characteristics of pesticides and the classification of agricultural products which organized into a hierarchical structure are complex and vary widely across different regions or countries. So it is a big challenge to compare multi-regional MRL standard data comprehensively. In this paper, we present a multi-comparable visual analytic approach for complex hierarchical data and a visual analytics system (McVA) to support multiple comparison and evaluation of MRL standard. With a cooperative multi-view visual design, our proposed approach links the hierarchies of MRL datasets and provides the capacity for comparison at different levels and dimensions. We also introduce a metric model for evaluating the completeness and strictness of MRL standards quantitatively. The case study of real problems and the positive feedback from domain experts demonstrate the effectiveness of this approach.

A Multi-comparable visual analytic approach for complex hierarchical data

The invention discloses a comparison correlation visual analysis method (comparison relevance tree) aimed at two kinds of hierarchical data, and application. The method comprises: respectively using a node-link tree based on force orientation and a space filling tree based on radiation rings to represent two kinds of hierarchical data, nodes in the node-link tree representing as a pie chart, space filling of the radiation rings being distinguished by colors; and through an interactive method, showing comparison and correlation of different hierarchical data of the two kinds of data. The visual method can show hierarchical relations and association of two different kinds of hierarchical data at the same time, and can directly compare correlation information and other data of the same kind of hierarchical data, and is intuitive and effective, and visual disorder is prevented. The method helps users to efficiently check associations among data, so as to perform further data analysis.

Comparison correlation visual analysis method aimed at two kinds of hierarchical data, and application

According to the characteristics of pesticide residue data and analyzing requirements in food safety fields, we presented a visual analyzing method for associated hierarchical data, called sunburst with ordered nodes based on hierarchical clustering (SONHC). SONHC arranged the leaf nodes in sunburst in order using hierarchical clustering algorithm, put the associated dataset as a node in center of the sunburst, and connected it with the associated leaf nodes in sunburst using colored lines. So, it can present not only two hierarchical structures but also the relationships between them. Based on SONHC and some interaction techniques (clicking, contraction and expansion, etc) we developed an associated visual analyzing system (AVAS) for pesticide residues detection results data, which can help users to inspect the hierarchical structure of pesticide and agricultural products and to explore the associations between pesticides and agricultural products, and associations between different pesticides. The results of user experience test showed that SONHC algorithm overperforms than SA and SR algorithm in ULE and ULE’s variance. AVAS system is effective in helping users to analyze the pesticide residues data. Furthermore, SONHC algorithm can also be adopted to analyze associated hierarchical data in other fields, such as finance, insurance and e-commerce.

Sunburst with ordered nodes based on hierarchical clustering: a visual analyzing method for associated hierarchical pesticide residue data

The invention belongs to the field of information visualization, and particularly relates to a visualization method of two sorts of associated hierarchical data. The visualization method is used for visualizing the associated hierarchical data. According to the visualization method, two independent hierarchical data are visualized based on a force-oriented node-link tree and based on a space filling tree of a radiation ring respectively; then nodes with incidence relation in the two trees are connected through a straight line to form a double-relevance tree. In order to reduce edge crossing and visual confusion phenomena, an invisible ring concept is introduced through a path rerouting layout optimization algorithm, and a cubic bezeir curve is used for replacing a straight line to optimize a relation edge. According to the visualization method for the two sorts of associated hierarchical data, the hierarchical relation and the incidence relation in the two sorts of hierarchical data are showed at the same time, visual confusion is effectively reduced, a user is helped to check the incidence relation between the data, and further data analysis is facilitated.

Xinyue Zhang

Papers

Sunburst with ordered nodes based on hierarchical clustering: a visual analyzing method for associated hierarchical pesticide residue data

Visualization method of two sorts of associated hierarchical data