Density based fuzzy c-means clustering of non-convex patterns

We propose a new technique to perform unsupervised data classification (clustering) based on density induced metric and non-smooth optimization. Our goal is to automatically recognize multidimensional clusters of non-convex shape. We present a modification of the fuzzy c-means algorithm, which uses the data induced metric, defined with the help of Delaunay triangulation. We detail computation of the distances in such a metric using graph algorithms. To find optimal positions of cluster prototypes we employ the discrete gradient method of non-smooth optimization. The new clustering method is capable to identify non-convex overlapped d-dimensional clusters.

The density-based agglomerative information bottleneck

The Information Bottleneck method aims to extract a compact representation which preserves the maximum relevant information. The sub-optimality in agglomerative Information Bottleneck (aIB) algorithm restricts the applications of Information Bottleneck method. In this paper, the concept of density-based chains is adopted to evaluate the information loss among the neighbors of an element, rather than the information loss between pairs of elements. The DaIB algorithm is then presented to alleviate the sub-optimality problem in aIB while simultaneously keeping the useful hierarchical clustering tree-structure. The experiment results on the benchmark data sets show that the DaIB algorithm can get more relevant information and higher precision than aIB algorithm, and the paired t-test indicates that these improvements are statistically significant.

Automatic generation of traditional style painting by using density-based color clustering

In this paper, a novel approach is proposed to automatically generate both watercolor painting and pencil sketch drawing, or binary image of contour, from realism-style photo by using DBSCAN color clustering based on HSV color space. While the color clusters produced by proposed methods help to create watercolor painting, the noise pixels are useful to generate the pencil sketch drawing. Moreover, noise pixels are reassigned to color clusters by a novel algorithm to refine the contour in the watercolor painting. The main goal of this paper is to inspire non-professional artists' imagination to produce traditional style painting easily by only adjusting a few parameters. Also, another contribution of this paper is to propose an easy method to produce the binary image of contour, which is a vice product when mining image data by DBSCAN clustering. Thus the binary image is useful in resource limited system to reduce data but keep enough information of images. © 2007 IEEE.

Fuzzy c-means density based clustering using data induced metric

We propose a new data induced metric to perform un supervised data classification (clustering). Our goal is to automatically recognize clusters of non-convex shape. We present a new version of fuzzy c-means al gorithm, based on the data induced metric, which is capable to identify non-convex d-dimensional clusters.

Development and application of a microelectrode based scanning voltammetric detector

A large part of the work presented in this thesis describes the development and use of a novel electrochemical detector designed to allow the electrochemical characterisation of compounds in flowing solution by means of cyclic voltammetry. The detector was microprocessor controlled, which provides digital generation of the potential waveform and collection of data for subsequent analysis. Microdisk working electrodes are employed to permit both thermodynamic and kinetically controlled processes to be studied under steady-state conditions in flowing solutions without the distortion or hysteresis normally encountered with larger sized electrodes. The effect of electrode size, potential scan rate, and solution flow rate are studied extensively with the oxidation of ferrocene used as an example of a thermodynamically controlled process and a series of catecholamines as examples of a kinetically controlled process. The performance of the detector was best demonstrated when used as a HPLC post-column detector. The 3-dimensional chromatovoltammograms obtained allow on-line characterisation of each fraction as it elutes from the column. The rest of the work presented in this thesis involves the study of the oxidative degradation pathway of dithranol. The oxidative pathway was shown to involve a complex free radical mechanism, dependent on the presence of both oxygen and, in particular light. The pathway is further complicated by the fact that dithranol may exist in either a keto or enol form, the enol being most susceptible to oxidation. A likely mechanism is proposed from studies performed with cyclic voltammetry and controlled potential electrolysis, then defined by subsequent kinetic studies.

Candidate working set strategy based SMO algorithm in support vector machine

Sequential minimal optimization (SMO) is quite an efficient algorithm for training the support vector machine. The most important step of this algorithm is the selection of the working set, which greatly affects the training speed. The feasible direction strategy for the working set selection can decrease the objective function, however, may augment to the total calculation for selecting the working set in each of the iteration. In this paper, a new candidate working set (CWS) Strategy is presented considering the cost on the working set selection and cache performance. This new strategy can select several greatest violating samples from Cache as the iterative working sets for the next several optimizing steps, which can improve the efficiency of the kernel cache usage and reduce the computational cost related to the working set selection. The results of the theory analysis and experiments demonstrate that the proposed method can reduce the training time, especially on the large-scale datasets.

A new physarum network based genetic algorithm for bandwidth-delay constrained least-cost multicast routing

Bandwidth-delay constrained least-cost multicast routing is a typical NP-complete problem. Although some swarm-based intelligent algorithms (e.g., genetic algorithm (GA)) are proposed to solve this problem, the shortcomings of local search affect the computational effectiveness. Taking the ability of building a robust network of Physarum network model (PN), a new hybrid algorithm, Physarum network-based genetic algorithm (named as PNGA), is proposed in this paper. In PNGA, an updating strategy based on PN is used for improving the crossover operator of traditional GA, in which the same parts of parent chromosomes are reserved and the new offspring by the Physarum network model is generated. In order to estimate the effectiveness of our proposed optimized strategy, some typical genetic algorithms and the proposed PNGA are compared for solving multicast routing. The experiments show that PNGA has more efficient than original GA. More importantly, the PNGA is more robustness that is very important for solving the multicast routing problem.

The impact of a line probe assay based diagnostic algorithm on time to treatment initiation and treatment outcomes for multidrug resistant TB patients in Arkhangelsk region, Russia

BACKGROUND: In the Arkhangelsk region of Northern Russia, multidrug-resistant (MDR) tuberculosis (TB) rates in new cases are amongst the highest in the world. In 2014, MDR-TB rates reached 31.7% among new cases and 56.9% among retreatment cases. The development of new diagnostic tools allows for faster detection of both TB and MDR-TB and should lead to reduced transmission by earlier initiation of anti-TB therapy. STUDY AIM: The PROVE-IT (Policy Relevant Outcomes from Validating Evidence on Impact) Russia study aimed to assess the impact of the implementation of line probe assay (LPA) as part of an LPA-based diagnostic algorithm for patients with presumptive MDR-TB focusing on time to treatment initiation with time from first-care seeking visit to the initiation of MDR-TB treatment rather than diagnostic accuracy as the primary outcome, and to assess treatment outcomes. We hypothesized that the implementation of LPA would result in faster time to treatment initiation and better treatment outcomes.

METHODS: A culture-based diagnostic algorithm used prior to LPA implementation was compared to an LPA-based algorithm that replaced BacTAlert and Löwenstein Jensen (LJ) for drug sensitivity testing. A total of 295 MDR-TB patients were included in the study, 163 diagnosed with the culture-based algorithm, 132 with the LPA-based algorithm.

RESULTS: Among smear positive patients, the implementation of the LPA-based algorithm was associated with a median decrease in time to MDR-TB treatment initiation of 50 and 66 days compared to the culture-based algorithm (BacTAlert and LJ respectively, p<0.001). In smear negative patients, the LPA-based algorithm was associated with a median decrease in time to MDR-TB treatment initiation of 78 days when compared to the culture-based algorithm (LJ, p<0.001). However, several weeks were still needed for treatment initiation in LPA-based algorithm, 24 days in smear positive, and 62 days in smear negative patients. Overall treatment outcomes were better in LPA-based algorithm compared to culture-based algorithm (p = 0.003). Treatment success rates at 20 months of treatment were higher in patients diagnosed with the LPA-based algorithm (65.2%) as compared to those diagnosed with the culture-based algorithm (44.8%). Mortality was also lower in the LPA-based algorithm group (7.6%) compared to the culture-based algorithm group (15.9%). There was no statistically significant difference in smear and culture conversion rates between the two algorithms.

CONCLUSION: The results of the study suggest that the introduction of LPA leads to faster time to MDR diagnosis and earlier treatment initiation as well as better treatment outcomes for patients with MDR-TB. These findings also highlight the need for further improvements within the health system to reduce both patient and diagnostic delays to truly optimize the impact of new, rapid diagnostics.

A novel time computation model based on algorithm complexity for data intensive scientific workflow design and scheduling

Scientific workflow offers a framework for cooperation between remote and shared resources on a grid computing environment (GCE) for scientific discovery. One major function of scientific workflow is to schedule a collection of computational subtasks in well-defined orders for efficient outputs by estimating task duration at runtime. In this paper, we propose a novel time computation model based on algorithm complexity (termed as TCMAC model) for high-level data intensive scientific workflow design. The proposed model schedules the subtasks based on their durations and the complexities of participant algorithms. Characterized by utilization of task duration computation function for time efficiency, the TCMAC model has three features for a full-aspect scientific workflow including both dataflow and control-flow: (1) provides flexible and reusable task duration functions in GCE;(2) facilitates better parallelism in iteration structures for providing more precise task durations;and (3) accommodates dynamic task durations for rescheduling in selective structures of control flow. We will also present theories and examples in scientific workflows to show the efficiency of the TCMAC model, especially for control-flow. Copyright©2009 John Wiley & Sons, Ltd.

A C-DBSCAN algorithm for determining bus-stop locations based on taxi GPS data

Determining suitable bus-stop locations is critical in improving the quality of bus services. Previous studies on selecting bus stop locations mainly consider environmental factors such as population density and traffic conditions, seldom of them consider the travel patterns of people, which is a key factor for determining bus-stop locations. In order to draw people’s travel patterns, this paper improves the density-based spatial clustering of applications with noise (DBSCAN) algorithm to find hot pick-up and drop-off locations based on taxi GPS data. The discovered density-based hot locations could be regarded as the candidate for bus-stop locations. This paper further utilizes the improved DBSCAN algorithm, namely as C-DBSCAN in this paper, to discover candidate bus-stop locations to Capital International Airport in Beijing based on taxi GPS data in November 2012. Finally, this paper discusses the effects of key parameters in C-DBSCAN algorithm on the clustering results. Keywords Bus-stop locations, Public transport service, Taxi GPS data, Centralize density-based spatial clustering of applications with noise.

An adaptive policy-based vertical handoff algorithm for heterogeneous wireless networks

The next generation of wireless networks is envisioned as convergence of heterogeneous radio access networks. Since technologies are becoming more collaborative, a possible integration between IEEE 802.16 based network and previous generation of telecommunication systems (2G, ..., 3G) must be considered. A novel quality function based vertical handoff (VHO) algorithm, based on proposed velocity and average receive power estimation algorithms is discussed in this paper. The short-time Fourier analysis of received signal strength (RSS) is employed to obtain mobile speed and average received power estimates. Performance of quality function based VHO algorithm is evaluated by means of measure of quality of service (QoS). Simulation results show that proposed quality function, brings significant gains in QoS and more efficient use of resources can be achieved.