Particle Competition and Cooperation in Networks for Semi-Supervised Learning

Semi-supervised learning is one of the important topics in machine learning, concerning with pattern classification where only a small subset of data is labeled. In this paper, a new network-based (or graph-based) semi-supervised classification model is proposed. It employs a combined random-greedy walk of particles, with competition and cooperation mechanisms, to propagate class labels to the whole network. Due to the competition mechanism, the proposed model has a local label spreading fashion, i.e., each particle only visits a portion of nodes potentially belonging to it, while it is not allowed to visit those nodes definitely occupied by particles of other classes. In this way, a "divide-and-conquer" effect is naturally embedded in the model. As a result, the proposed model can achieve a good classification rate while exhibiting low computational complexity order in comparison to other network-based semi-supervised algorithms. Computer simulations carried out for synthetic and real-world data sets provide a numeric quantification of the performance of the method.

Thrombin-Receptor Antagonist Vorapaxar in Acute Coronary Syndromes

BACKGROUND Vorapaxar is a new oral protease-activated receptor 1 (PAR-1) antagonist that inhibits thrombin-induced platelet activation. METHODS In this multinational, double-blind, randomized trial, we compared vorapaxar with placebo in 12,944 patients who had acute coronary syndromes without ST-segment elevation. The primary end point was a composite of death from cardiovascular causes, myocardial infarction, stroke, recurrent ischemia with rehospitalization, or urgent coronary revascularization. RESULTS Follow-up in the trial was terminated early after a safety review. After a median follow-up of 502 days (interquartile range, 349 to 667), the primary end point occurred in 1031 of 6473 patients receiving vorapaxar versus 1102 of 6471 patients receiving placebo (Kaplan-Meier 2-year rate, 18.5010 vs. 19.9%; hazard ratio, 0.92; 95% confidence interval [CI], 0.85 to 1.01; P=0.07). A composite of death from cardiovascular causes, myocardial infarction, or stroke occurred in 822 patients in the vorapaxar group versus 910 in the placebo group (14.7% and 16.4%, respectively; hazard ratio, 0.89; 95% CI, 0.81 to 0.98; P=0.02). Rates of moderate and severe bleeding were 7.2% in the vorapaxar group and 5.2% in the placebo group (hazard ratio, 1.35; 95% CI, 1.16 to 1.58; P<0.001). Intracranial hemorrhage rates were 1.1% and 0.2%, respectively (hazard ratio, 3.39; 95% CI, 1.78 to 6.45; P<0.001). Rates of nonhemorrhagic adverse events were similar in the two groups. CONCLUSIONS In patients with acute coronary syndromes, the addition of vorapaxar to standard therapy did not significantly reduce the primary composite end point but significantly increased the risk of major bleeding, including intracranial hemorrhage. (Funded by Merck; TRACER ClinicalTrials.gov number, NCT00527943.)

Collaborative fuzzy clustering algorithms: some refinements and design guidelines

There are some variants of the widely used Fuzzy C-Means (FCM) algorithm that support clustering data distributed across different sites. Those methods have been studied under different names, like collaborative and parallel fuzzy clustering. In this study, we offer some augmentation of the two FCM-based clustering algorithms used to cluster distributed data by arriving at some constructive ways of determining essential parameters of the algorithms (including the number of clusters) and forming a set of systematically structured guidelines such as a selection of the specific algorithm depending on the nature of the data environment and the assumptions being made about the number of clusters. A thorough complexity analysis, including space, time, and communication aspects, is reported. A series of detailed numeric experiments is used to illustrate the main ideas discussed in the study.

Prasugrel versus Clopidogrel for Acute Coronary Syndromes without Revascularization

Background The effect of intensified platelet inhibition for patients with unstable angina or myocardial infarction without ST-segment elevation who do not undergo revascularization has not been delineated. Methods In this double-blind, randomized trial, in a primary analysis involving 7243 patients under the age of 75 years receiving aspirin, we evaluated up to 30 months of treatment with prasugrel (10 mg daily) versus clopidogrel (75 mg daily). In a secondary analysis involving 2083 patients 75 years of age or older, we evaluated 5 mg of prasugrel versus 75 mg of clopidogrel. Results At a median follow-up of 17 months, the primary end point of death from cardiovascular causes, myocardial infarction, or stroke among patients under the age of 75 years occurred in 13.9% of the prasugrel group and 16.0% of the clopidogrel group (hazard ratio in the prasugrel group, 0.91; 95% confidence interval [CI], 0.79 to 1.05; P = 0.21). Similar results were observed in the overall population. The prespecified analysis of multiple recurrent ischemic events (all components of the primary end point) suggested a lower risk for prasugrel among patients under the age of 75 years (hazard ratio, 0.85; 95% CI, 0.72 to 1.00; P = 0.04). Rates of severe and intracranial bleeding were similar in the two groups in all age groups. There was no significant between-group difference in the frequency of nonhemorrhagic serious adverse events, except for a higher frequency of heart failure in the clopidogrel group. Conclusions Among patients with unstable angina or myocardial infarction without ST- segment elevation, prasugrel did not significantly reduce the frequency of the primary end point, as compared with clopidogrel, and similar risks of bleeding were observed. (Funded by Eli Lilly and Daiichi Sankyo; TRILOGY ACS ClinicalTrials.gov number, NCT00699998.)

Prediction of enzymatic infarct size in ST-segment elevation myocardial infarction

Objectives Predictors of adverse outcomes following myocardial infarction (MI) are well established; however, little is known about what predicts enzymatically estimated infarct size in patients with acute ST-elevation MI. The Complement And Reduction of INfarct size after Angioplasty or Lytics trials of pexelizumab used creatine kinase (CK)-MB area under the curve to determine infarct size in patients treated with primary percutaneous coronary intervention (PCI) or fibrinolysis. Methods Prediction of infarct size was carried out by measuring CK-MB area under the curve in patients with ST-segment elevation MI treated with reperfusion therapy from January 2000 to April 2002. Infarct size was calculated in 1622 patients (PCI=817; fibrinolysis=805). Logistic regression was used to examine the relationship between baseline demographics, total ST-segment elevation, index angiographic findings (PCI group), and binary outcome of CK-MB area under the curve greater than 3000 ng/ml. Results Large infarcts occurred in 63% (515) of the PCI group and 69% (554) of the fibrinolysis group. Independent predictors of large infarcts differed depending on mode of reperfusion. In PCI, male sex, no prior coronary revascularization and diabetes, decreased systolic blood pressure, sum of ST-segment elevation, total (angiographic) occlusion, and nonright coronary artery culprit artery were independent predictors of larger infarcts (C index=0.73). In fibrinolysis, younger age, decreased heart rate, white race, no history of arrhythmia, increased time to fibrinolytic therapy in patients treated up to 2 h after symptom onset, and sum of ST-segment elevation were independently associated with a larger infarct size (C index=0.68). Conclusion Clinical and patient data can be used to predict larger infarcts on the basis of CK-MB quantification. These models may be helpful in designing future trials and in guiding the use of novel pharmacotherapies aimed at limiting infarct size in clinical practice. Coron Artery Dis 23:118-125 (C) 2012 Wolters Kluwer Health | Lippincott Williams & Wilkins.