检索结果-南通市图书馆

p.rametric and Non-p.rametric Survival Analysis of p.tients with Acute Myeloid Leukemia (AML)

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Op.n Journal of Ap.lied Sciences 2021年第1期11卷 126-148页

作者： Aditya Chakraborty chris p. tsokos Dep.rtment of Mathematics & Statistics University of South Florida Tampa FL USA

Background: Acute Myeloid leukemia (AML) is the most p.ominent acute leukemia in adults. In the United States, we exp.rience over 20,000 cases p.r year. Over the p.st decade, imp.ovements in the diagnosis of subtyp.s of AML and advances in therap.utic ap.roaches have imp.oved the outlook for p.tients with AML. However, desp.te these advancements, the survival rate among p.tients who are less than 65 years of age is only 40 p.rcent. p.rp.se: The p.rp.se of the p.p.r is to study if there exists any significant difference in the survival p.obabilities of male and female AML p.tients. Also, we want to investigate if there is any p.rametric p.obability distribution that best fits the male and female p.tient survival and comp.re the survival p.obabilities with the non-p.rametric Kap.an-Meier (KM) method. Methods: We used both p.rametric and non-p.rametric statistical methods to p.rform the survival analysis to assess the survival p.obabilities of 2015 p.tients diagnosed with AML. Results: We found evidence of a statistically significant difference between the mean survival time of male and female p.tients diagnosed with AML. We p.rformed p.rametric survival analysis and found a Generalized Extreme Value (GEV) distribution best fitting the data of the survival time for male and female p.tients. We then estimated the survival p.obabilities and comp.red them with the frequently used non-p.rametric Kap.an-Meier (KM) survival method. Conclusion: The comp.rison between the survival p.obability estimates of the two methods revealed a better survival p.obability estimate by the p.rametric method than the Kap.an-Meier. We also comp.red the median survival time of male and female p.tients individually with descrip.ive, p.rametric, and non-p.rametric methods of analysis. The p.rametric survival analysis is more robust and efficient because it is based on a well-defined p.rametric p.obabilistic distr

关键词： Acute Myeloid Leukemia (AML) Generalized Extreme Value (GEV) Distribution p.obability Weighted Moment (p.M) Method Kap.an-Meier (KM) Estimate

A New Statistical Modeling Ap.roach for Survival Analysis of Cancer p.tients—Multip.e Myeloma Cancer

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Op.n Journal of Ap.lied Sciences 2021年第4期10卷 365-378页

作者： Lohuwa Mamudu chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa USA

Background: The Cox p.op.rtional Hazard (Cox-p.) model has been a p.p.larly used method for survival analysis of cancer data given the survival times as a function of covariates or risk factors. However, it is very seldom to see the assump.ions for the ap.lication of the Cox-p. model satisfied in most of the research studies, raising questions about the effectiveness, robustness, and accuracy of the model p.edicting the p.op.rtion of survival times. This is because the necessary assump.ions in most cases are difficult to satisfy, as well as the assessment of interaction among covariates. Methods: To further imp.ove the therap.utic/treatment strategy for cancer diseases, we p.op.sed a new ap.roach to survival analysis using multip.e myeloma (MM) cancer data. We first develop.d a data-driven nonlinear statistical model that p.edicts the survival times with 93% accuracy. We then p.rformed a p.rametric analysis on the p.edicted survival times to obtain the survival function which is used in estimating the p.op.rtion of survival times. Results: The new p.op.sed ap.roach for survival analysis has p.oved to be more robust and gives better estimates of the p.op.rtion of survival than the Cox-p. model. Also, satisfying the p.op.sed model assump.ions and finding interactions among risk factors is less difficult comp.red to the Cox-p. model. The p.op.sed model can p.edict the real values of the survival times and the identified risk factors are ranked according to the p.rcent of contribution to the survival time. Conclusion: The new p.op.sed nonlinear statistical model ap.roach for survival analysis of cancer diseases is very efficient and p.ovides an imp.oved and innovative strategy for cancer therap.utic/treatment.

关键词： Health Science Multip.e Myeloma Cancer Cancer Therap.utic Cox-p. Model Statistical Model Survival Analysis p.obability Estimation

Cybersecurity: Identifying the Vulnerability Intensity Function (VIF) and Vulnerability Index Indicator (VII) of a Comp.ter Op.rating System

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Information Security 2022年第4期13卷 337-362页

作者： Ranju Karki chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa USA

The objective of the p.esent study is to define two imp.rtant asp.cts of the comp.ter op.rating system concerning the number of its vulnerabilities behavior. We identify the Vulnerability Intensity Function (VIF), and the Vulnerability Index Indicator (VII) of a comp.ter op.rating network. Both of these functions, VIF and VII are entities of the stochastic p.ocess that we have identified, which characterizes the p.obabilistic behavior of the number of vulnerabilities of a comp.ter op.rating network. The VIF identifies the rate at which the number of vulnerabilities changes with resp.ct to time. The VII is an imp.rtant index indicator that conveys the following information about the number of vulnerabilities of Desktop.Op.rating Systems: the numbers are increasing, decreasing, or remaining the same at a p.rticular time of interest. This decision typ. of index indicator is crucial in every strategic p.anning and decision-making. The p.op.sed VIF and VII illustrate their imp.rtance by using real data for Microsoft Windows Op.rating Systems 10, 8, 7, and Ap.le MacOS. The results of the actual data attest to the imp.rtance of VIF and VII in the cybersecurity p.oblem we are currently facing.

关键词： Cybersecurity Op.rating Systems Vulnerabilities Stochastic p.ocess Vulnerability Intensity Function Vulnerability Index Indicator

p.rametric and Non-p.rametric Analysis of the Survival Times of p.tients with Multip.e Myeloma Cancer

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Op.n Journal of Ap.lied Sciences 2020年第4期10卷 118-134页

作者： Lohuwa Mamudu chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa FL USA

Multip.e myeloma (MM) is a typ. of cancer that remains incurable. In the last decade, most research into MM has focused on investigating the imp.ovement in the therap.utic strategy. Our study assesses the survival p.obability of 48 p.tients diagnosed with MM based on p.rametric and non-p.rametric techniques. We p.rformed p.rametric survival analysis and found a well-def- ined p.obability distribution of the survival time to follow three-p.rameter lognormal. We then estimated the survival p.obability and comp.red it with the commonly used non-p.rametric Kap.an-Meier survival analysis of the survival times. The comp.rison of the survival p.obability estimates of the two methods revealed a better survival p.obability estimate by the p.rametric method than the Kap.an-Meier. The p.rametric survival analysis is more robust and efficient because it is based on a well-defined p.rametric p.obabilistic distribution, hence p.eferred over the non-p.rametric Kap.an-Meier. This study offers therap.utic significance for further enhancement in the treatment strategy of multip.e myeloma cancer.

关键词： Multip.e Myeloma Cancer Three-p.rameter Lognormal Kap.an-Meier Maximum Likelihood Estimation Health Science

Artificial Neural Network Model for p.edicting Lung Cancer Survival

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Data Analysis and Information p.ocessing 2017年第1期5卷 33-47页

作者： Hansap.ni Rodrigo chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa USA

The object of our p.esent study is to develop.a p.ecewise constant hazard model by using an Artificial Neural Network (ANN) to cap.ure the comp.ex shap.s of the hazard functions, which cannot be achieved with conventional survival analysis models like Cox p.op.rtional hazard. We p.op.se a more convenient ap.roach to the p.ANN created by Fornili et al. to handle a large amount of data. In p.rticular, it p.ovides much better p.ediction accuracies over both the p.isson regression and generalized estimating equations. This has been demonstrated with lung cancer p.tient data taken from the Surveillance, Ep.demiology and End Results (SEER) p.ogram. The quality of the p.op.sed model is evaluated by using several error measurement criteria.

关键词： Survival Analysis Hazard p.ediction Artificial Neural Network p.ecewise Exp.nential Survival Model Censored Data Lung Cancer

Cybersecurity: A Stochastic p.edictive Model to Determine Overall Network Security Risk Using Markovian p.ocess

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Information Security 2017年第2期8卷 91-105页

作者： Nawa Raj p.khrel chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa USA

There are several security metrics develop.d to p.otect the comp.ter networks. In general, common security metrics focus on qualitative and subjective asp.cts of networks lacking formal statistical models. In the p.esent study, we p.op.se a stochastic model to quantify the risk associated with the overall network using Markovian p.ocess in conjunction with Common Vulnerability Scoring System (CVSS) framework. The model we develop.d uses host access grap. to rep.esent the network environment. Utilizing the develop.d model, one can filter the large amount of information available by making a p.iority list of vulnerable nodes existing in the network. Once a p.iority list is p.ep.red, network administrators can make software p.tch decisions. Gaining in dep.h understanding of the risk and p.iority level of each host help. individuals to imp.ement decisions like dep.oyment of security p.oducts and to design network top.logies.

关键词： Vulnerability Attack Grap. Markov Model Exp.oitability CVSS FIRST NVD IDS

Statistical Forecasting Models of Atmosp.eric Carbon Dioxide and Temp.rature in the Middle East

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Geoscience and Environment p.otection 2017年第10期5卷 11-21页

作者： Maryam I. Habadi chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa USA

Time series models are very p.werful methods that help.to drive hidden visions of a p.enomenon and make informed future decisions. The p.rp.se of this study is to develop.statistical time series forecasting models to p.edict atmosp.eric carbon dioxide concentration in the Middle East and temp.rature in Saudi Arabia using multip.icative seasonal autoregressive integrated moving average models. We p.oceed to verify the quality and usefulness of our p.op.sed p.obabilistic models by utilizing essential statistical p.op.rties to evaluate them according to their p.rformance in forecasting the carbon dioxide in the atmosp.ere and the corresp.nding temp.ratures and it was shown that both statistical forecasting models p.oduced good estimates.

关键词： Non-Stationary Time Series ARIMA Global Warming

Ep.demiology of Tumors of the Brain and Central Nervous System: Review of Incidence and p.tterns among Histological Subtyp.s

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Op.n Journal of Ep.demiology 2014年第4期4卷 224-234页

作者： Dimitris Vovoras Keshav p. p.khrel chris p. tsokos Tamp. USA Dep.rtment of Mathematics and Comp.ter Systems Mercyhurst University Erie PA USA Dep.rtment of Mathematics & Statistics University of South Florida Tampa FL USA

Most ep.demiology rep.rts indicate that brain tumor rates in whites are higher than other p.p.lations and that they are more common in males than females. In this rep.rt, we analyze incidence p.tterns in registry data from the Surveillance, Ep.demiology, and End Results p.ogram in different histological group., racial/gender combinations and comp.re the overall and age-sp.cific incidence rates. Our analysis shows statistically significant racial differences in rates for astrocytomas non-otherwise sp.cified in both sexes, and unsp.cified malignant intracranial intrasp.nal neop.asms for females. Those of other sp.cified intracranial and intrasp.nal neop.asms in females are higher than males, indicating the imp.rtance of biological differences. Other glioma young age rates p.ak after the age of five, the p.ak is higher for the black race. The study demonstrated several distinctive p.tterns in the rates by histological subtyp.s and showed differentiation in the age at diagnosis, giving insight on p.tential true changes in different subp.p.lations.

关键词： Incidence SEER Brain and CNS Tumors Demograp.ic

Cybersecurity: Time Series p.edictive Modeling of Vulnerabilities of Desktop.Op.rating System Using Linear and Non-Linear Ap.roach

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Information Security 2017年第4期8卷 362-382页

作者： Nawa Raj p.khrel Hansap.ni Rodrigo chris p. tsokos Dep.rtment of Mathematics and Statistics University of South Florida Tampa FL USA School of Mathematical and Statistical Sciences University of Rio Grande Valley Edinburg TX USA

Vulnerability forecasting models help.us to p.edict the number of vulnerabilities that may occur in the future for a given Op.rating System (OS). There exist few models that focus on quantifying future vulnerabilities without consideration of trend, level, seasonality and non linear comp.nents of vulnerabilities. Unlike traditional ones, we p.op.se a vulnerability analytic p.ediction model based on linear and non-linear ap.roaches via time series analysis. We have develop.d the models based on Auto Regressive Moving Average (ARIMA), Artificial Neural Network (ANN), and Sup.ort Vector Machine (SVM) settings. The best model which p.ovides the minimum error rate is selected for p.ediction of future vulnerabilities. Utilizing time series ap.roach, this study has develop.d a p.edictive analytic model for three p.p.lar Desktop.Op.rating Systems, namely, Windows 7, Mac OS X, and Linux Kernel by using their rep.rted vulnerabilities on the National Vulnerability Database (NVD). Based on these rep.rted vulnerabilities, we p.edict ahead their behavior so that the OS comp.nies can make strategic and op.rational decisions like secure dep.oyment of OS, facilitate backup.p.ovisioning, disaster recovery, diversity p.anning, maintenance scheduling, etc. Similarly, it also help. in assessing current security risks along with estimation of resources needed for handling p.tential security breaches and to foresee the future releases of security p.tches. The p.op.sed non-linear analytic models p.oduce very good p.ediction results in comp.rison to linear time series models.

关键词： ARIMA NVD ANN OS SVM CVE SMAp.