Connectivity-based Clustering Research Articles

Connectivity-Based Clustering for Mixed Discrete and Continuous Data

This paper introduces a density-based clustering procedure for datasets with variables of mixed type. The proposed procedure, which is closely related to the concept of shared neighbourhoods, works particularly well in cases where the individual clusters differ greatly in terms of the average pairwise distance of the associated objects. Using a number of concrete examples, it is shown that the proposed clustering algorithm succeeds in allowing the identification of subgroups of objects with statistically significant distributional characteristics.

Facial expression recognition is linked to clinical and neurofunctional differences in autism

BackgroundDifficulties in social communication are a defining clinical feature of autism. However, the underlying neurobiological heterogeneity has impeded targeted therapies and requires new approaches to identifying clinically relevant bio-behavioural subgroups. In the largest autism cohort to date, we comprehensively examined difficulties in facial expression recognition, a key process in social communication, as a bio-behavioural stratification biomarker, and validated them against clinical features and neurofunctional responses.MethodsBetween 255 and 488 participants aged 6–30 years with autism, typical development and/or mild intellectual disability completed the Karolinska Directed Emotional Faces task, the Reading the Mind in the Eyes Task and/or the Films Expression Task. We first examined mean-group differences on each test. Then, we used a novel intersection approach that compares two centroid and connectivity-based clustering methods to derive subgroups based on the combined performance across the three tasks. Measures and subgroups were then related to clinical features and neurofunctional differences measured using fMRI during a fearful face-matching task.ResultsWe found significant mean-group differences on each expression recognition test. However, cluster analyses showed that these were driven by a low-performing autistic subgroup (~ 30% of autistic individuals who performed below 2SDs of the neurotypical mean on at least one test), while a larger subgroup (~ 70%) performed within 1SD on at least 2 tests. The low-performing subgroup also had on average significantly more social communication difficulties and lower activation in the amygdala and fusiform gyrus than the high-performing subgroup.LimitationsFindings of autism expression recognition subgroups and their characteristics require independent replication. This is currently not possible, as there is no other existing dataset that includes all relevant measures. However, we demonstrated high internal robustness (91.6%) of findings between two clustering methods with fundamentally different assumptions, which is a critical pre-condition for independent replication.ConclusionsWe identified a subgroup of autistic individuals with expression recognition difficulties and showed that this related to clinical and neurobiological characteristics. If replicated, expression recognition may serve as bio-behavioural stratification biomarker and aid in the development of targeted interventions for a subgroup of autistic individuals.

Clustering by measuring local direction centrality for data with heterogeneous density and weak connectivity

Clustering is a powerful machine learning method for discovering similar patterns according to the proximity of elements in feature space. It is widely used in computer science, bioscience, geoscience, and economics. Although the state-of-the-art partition-based and connectivity-based clustering methods have been developed, weak connectivity and heterogeneous density in data impede their effectiveness. In this work, we propose a boundary-seeking Clustering algorithm using the local Direction Centrality (CDC). It adopts a density-independent metric based on the distribution of K-nearest neighbors (KNNs) to distinguish between internal and boundary points. The boundary points generate enclosed cages to bind the connections of internal points, thereby preventing cross-cluster connections and separating weakly-connected clusters. We demonstrate the validity of CDC by detecting complex structured clusters in challenging synthetic datasets, identifying cell types from single-cell RNA sequencing (scRNA-seq) and mass cytometry (CyTOF) data, recognizing speakers on voice corpuses, and testifying on various types of real-world benchmarks.

Drosophila gustatory projections are segregated by taste modality and connectivity.

Gustatory sensory neurons detect caloric and harmful compounds in potential food and convey this information to the brain to inform feeding decisions. To examine the signals that gustatory neurons transmit and receive, we reconstructed gustatory axons and their synaptic sites in the adult Drosophila melanogaster brain, utilizing a whole-brain electron microscopy volume. We reconstructed 87 gustatory projections from the proboscis labellum in the right hemisphere and 57 from the left, representing the majority of labellar gustatory axons. Gustatory neurons contain a nearly equal number of interspersed pre- and postsynaptic sites, with extensive synaptic connectivity among gustatory axons. Morphology- and connectivity-based clustering revealed six distinct groups, likely representing neurons recognizing different taste modalities. The vast majority of synaptic connections are between neurons of the same group. This study resolves the anatomy of labellar gustatory projections, reveals that gustatory projections are segregated based on taste modality, and uncovers synaptic connections that may alter the transmission of gustatory signals.

Gravity and depth of social media networks

Abstract Structures of social media networks provide a composite view of dyadic connectivity across social actors, which reveals the spread of local and global influences of those actors in the network. Although social media network is a construct inferred from online activities, an underlying feature is that the actors also possess physical locational characteristics. Using a unique dataset from Facebook that provides a snapshot of the complete enumeration of county-to-county connectivity in the USA (in April 2016), we exploit these two dimensions viz. online connectivity and geographic distance between the counties, to establish a mapping between the two. We document two major results. First, social connectivity wanes as physical distance increases between county-pairs, signifying gravity-like behaviour found in economic activities like trade and migration. Two, a geometric projection of the network on a lower-dimensional space allows us to quantify depth of the nodes in the network with a well-defined metric. Clustering of this projected network reveals that the counties belonging to the same cluster tend to exhibit geographic proximity, a finding we quantify with regression-based analysis as well. Thus, our analysis of the social media networks demonstrates a unique relationship between physical spatial clustering and node connectivity-based clustering. Our work provides a novel characterization of geometric distance in the study of social network analysis, linking abstract network topology with its statistical properties.

A Connectivity-Based Clustering Scheme for Intelligent Vehicles

The reliability, scalability, and stability of routing schemes are open challenges in highly evolving vehicular ad hoc networks (VANETs). Cluster-based routing is an efficient solution to cope with the dynamic and inconsistent structure of VANETs. In this paper, we propose a cluster-based routing scheme (hereinafter referred to as connectivity-based clustering), where link connectivity is used as a metric for cluster formation and cluster head (CH) selection. Link connectivity is a function of vehicle density and transmission range in the proposed connectivity-based clustering scheme. Moreover, we used a heuristic approach of spectral clustering for the optimal number of cluster formation. Lastly, an appropriate vehicle is selected as a CH based on the maximum Eigen-centrality score. The simulation results show that the suggested connectivity-based clustering scheme performs well in the optimal number of cluster selections, strongly connected (STC) route selection, and route request messages (RRMs) in the discovery of a particular path to the destination. Thus, we conclude that link connectivity and the heuristic approach of spectral clustering are valuable additions to existing routing schemes for high evolving networks.

LUET: A novel lines-of-uniformity based clustering protocol for heterogeneous-WSN for multiple-applications

Energy-efficient resource-optimization for multiple application scenarios requires appropriate clustering-based network-fractioning that skillfully removes unbalanced network connectivity, which is a stimulating bottleneck in WSN-performance. This paper proposes a new Uniform Connectivity-based clustering Protocol, called Lines-of-Uniformity based Enhanced-Threshold (LUET) to provide energy-efficient coverage in three-tier heterogeneous WSN. This protocol considers the node’s remnant-energy and its proximity from either of the lines of uniformity for lowering down average isolated node count in WSN. This paper also proposes a rotation epoch-based LUET variant, (LUET|R) that incorporates static epoch for initial clustering rounds until First-Node-Death to overcome the rapid fall after the death of the first node. The simulation model demonstrates the superiority of LUET and its variants over other established clustering protocols in terms of network-lifetime, power-efficiency, net death-rate, average isolated-nodes, throughput.

Machine Learning Ranks ECG as an Optimal Wearable Biosignal for Assessing Driving Stress

The demand for wearable devices that can detect anxiety and stress when driving is increasing. Recent studies have attempted to use multiple biosignals to detect driving stress. However, collecting multiple biosignals can be complex and is associated with numerous challenges. Determining the optimal biosignal for assessing driving stress can save lives. To the best of our knowledge, no study has investigated both longitudinal and transitional stress assessment using supervised and unsupervised ML techniques. Thus, this study hypothesizes that the optimal signal for assessing driving stress will consistently detect stress using supervised and unsupervised machine learning (ML) techniques. Two different approaches were used to assess driving stress: longitudinal (a combined repeated measurement of the same biosignals over three driving states) and transitional (switching from state to state such as city to highway driving). The longitudinal analysis did not involve a feature extraction phase while the transitional analysis involved a feature extraction phase. The longitudinal analysis consists of a novel interaction ensemble (INTENSE) that aggregates three unsupervised ML approaches: interaction principal component analysis, connectivity-based clustering, and K-means clustering. INTENSE was developed to uncover new knowledge by revealing the strongest correlation between the biosignal and driving stress marker. These three MLs each have their well-known and distinctive geometrical basis. Thus, the aggregation of their result would provide a more robust examination of the simultaneous non-causal associations between six biosignals: electrocardiogram (ECG), electromyogram, hand galvanic skin resistance, foot galvanic skin resistance, heart rate, respiration, and the driving stress marker. INTENSE indicates that ECG is highly correlated with the driving stress marker. The supervised ML algorithms confirmed that ECG is the most informative biosignal for detecting driving stress, with an overall accuracy of 75.02%.

Functional Brain Parcellations of the Infant Brain and the Associated Developmental Trends.

Resting-state functional connectivity studies have dramatically improved our understanding of the early human brain functional development during the past decade. However, one emerging problem that could potentially impede future progresses in the field is the definition of regions of interest (ROI), since it is well known that functional connectivity estimation can be seriously contaminated by within-ROI signal heterogeneity. In this study, based on a large-scale rsfMRI data set in human infants (230 neonates, 143 1-year olds, and 107 2-year olds), we aimed to derive a set of anatomically constrained, infant-specific functional brain parcellations using functional connectivity-based clustering. Our results revealed significantly higher levels of signal homogeneity within the newly defined functional parcellations compared with other schemes. Importantly, the global functional connectivity patterns associated with the newly defined functional subunits demonstrated significantly increasing levels of differentiation with age, confirming increasing levels of local specialization. Subsequent whole brain connectivity analysis revealed intriguing patterns of regional-level functional connectivity developments and system-level hub redistribution during infancy. Overall, the newly derived infant-specific functional brain parcellations and the associated novel developmental patterns will likely prove valuable for future early developmental studies using the functional connectivity technique.

Node Detection Using High-Dimensional Fuzzy Parcellation Applied to the Insular Cortex.

Several functional connectivity approaches require the definition of a set of regions of interest (ROIs) that act as network nodes. Different methods have been developed to define these nodes and to derive their functional and effective connections, most of which are rather complex. Here we aim to propose a relatively simple “one-step” border detection and ROI estimation procedure employing the fuzzy c-mean clustering algorithm. To test this procedure and to explore insular connectivity beyond the two/three-region model currently proposed in the literature, we parcellated the insular cortex of 20 healthy right-handed volunteers scanned in a resting state. By employing a high-dimensional functional connectivity-based clustering process, we confirmed the two patterns of connectivity previously described. This method revealed a complex pattern of functional connectivity where the two previously detected insular clusters are subdivided into several other networks, some of which are not commonly associated with the insular cortex, such as the default mode network and parts of the dorsal attentional network. Furthermore, the detection of nodes was reliable, as demonstrated by the confirmative analysis performed on a replication group of subjects.

Detecting protein complexes using connectivity among nodes in a PPI Network

In this paper, we introduce a connectivity-based protein complex finding method to find dense and sparse complexes along with core and periphery proteins. We named this method CNCM (Connectivity-based Network Clustering Method) as it uses clustering coefficient and the connectivity among nodes to find complexes. This method also ensures detection of protein complexes which are overlapping in nature. CNCM has been validated using both gold standard benchmark datasets and hand-curated PPI networks from publicly available sources. We evaluate and compare the performance of CNCM with several protein complex finding methods and found our results highly satisfactory.

Hashing-based lookup service in mobile ad-hoc networks with multi-hop stretchable clustering

We propose a cluster-based hashing lookup service for mobile ad hoc networks (MANETs) which consists of a multi-hop stretching clustering scheme and a size-based hashing lookup service. As a solution to the k-hop minimum dominating set, our connectivity-based stretchable clustering (CSC) scheme aims to minimise the number of clusterheads which are responsible for maintaining the cluster backbone. Based on the cluster topology, we present a cluster-based hashing lookup service (CHLS) which eliminate the issues of the grid system. Our method uses a size-based hash function and allows multiple bindings between one node and lookup servers, in order to reduce query delay time. They are scattered evenly in the networks by the multiple anchor clusters distribution algorithm. The present study concludes with extensive experiments showing that the CSC scheme attains stability with low leaving rate and the CHLS scheme has low communication overhead with a reasonably small query delay time.

A Data Mining Approach to Creating Fundamental Traffic Flow Diagram

This paper investigates application of clustering techniques in partitioning traffic flow data to congested and free flow regimes. Clustering techniques identify the similarities and dissimilarities between data, and classify the data into groups with similar characteristics. Such techniques have been successfully used in market research, astronomy, psychiatry, and transportation. A framework is proposed for clustering traffic data based on fundamental traffic flow variables. Three types of clustering techniques are investigated: 1) connectivity-based clustering, 2) centroid-based clustering, and 3) distribution-based clustering. Specifically, hierarchical clustering, K-means clustering and general mixture model (GMM) were investigated.Traffic sensor data from three freeway bottleneck locations in two major U.S. metropolitan areas, St. Louis, Missouri, and Twin Cities, Minnesota, were used in the study. Various combinations of traffic variables were investigated for all three clustering techniques. The results indicated that the clustering is an effective way to partition traffic data into the free flow and congested flow regimes. Partitioned traffic data can be used to create fundamental traffic flow diagrams and macroscopic traffic stream models. Using speeds, or both speeds and occupancies as input variables produced the best clustering results. The performance of K-means and hierarchical clustering techniques were comparable to each other and they outperformed GMM clustering.

Aberrant functional network recruitment of posterior parietal cortex in turner syndrome

Turner syndrome is a genetic disorder caused by the complete or partial absence of an X chromosome in affected women. Individuals with TS show characteristic difficulties with executive functions, visual-spatial and mathematical cognition, with relatively intact verbal skills, and congruent abnormalities in structural development of the posterior parietal cortex (PPC). The functionally heterogeneous PPC has recently been investigated using connectivity-based clustering methods, which sub-divide a given region into clusters of voxels showing similar structural or functional connectivity to other brain regions. In the present study, we extended this method to compare connectivity-based clustering between groups and investigate whether functional networks differentially recruit the PPC in TS. To this end, we parcellated the PPC into sub-regions based on temporal correlations with other regions of the brain. fMRI data were collected from 15 girls with TS and 14 typically developing (TD) girls, aged 7-14, while they performed a visual-spatial task. Temporal correlations between voxels in the PPC and a set of seed regions were calculated, and the PPC divided into clusters of voxels showing similar connectivity. It was found that in general the PPC parcellates similarly in TS and TD girls, but that regions in bilateral inferior parietal lobules, and posterior right superior parietal lobule, were reliably recruited by different networks in TS relative to TD participants. These regions showed weaker correlation in TS with a set of regions involved in visual processing. These results suggest that abnormal development of visuospatial functional networks in TS may relate to the well documented cognitive difficulties in this disorder.

Lightweight and Distributed Connectivity-Based Clustering Derived from Schelling's Model

Lightweight and Distributed Connectivity-Based Clustering Derived from Schelling's Model

A connectivity based clustering algorithm with application to VLSI circuit partitioning

Circuit partitioning is a fundamental problem in very large-scale integration (VLSI) physical design automation. In this brief, we present a new connectivity-based clustering algorithm for VLSI circuit partitioning. The proposed clustering method focuses on capturing natural clusters in a circuit, i.e., the groups of cells that are highly interconnected in a circuit. Therefore, the proposed clustering method can reduce the size of large-scale partitioning problems without losing partitioning solution qualities. The performance of the proposed clustering algorithm is evaluated on a standard set of partitioning benchmarks-ISPD98 benchmark suite. The experimental results show that by applying the proposed clustering algorithm, the previously reported best partitioning solutions from state-of-the-art partitioners are further improved.