Research Article Real-Time Detection of Application-Layer DDoS Attack Using Time Series Analysis
|
|
|
- Albert Simpson
- 10 years ago
- Views:
Transcription
1 Cotrol Sciece ad Egieerig Volume 2013, Article ID , 6 pages Research Article Real-Time Detectio of Applicatio-Layer DDoS Attack Usig Time Series Aalysis Togguag Ni, Xiaoqig Gu, Hogyua Wag, ad Yu Li School of Iformatio Sciece ad Egieerig, Chagzhou Uiversity, Chagzhou , Chia Correspodece should be addressed to Hogyua Wag; [email protected] Received 7 Jue 2013; Accepted 25 August 2013 Academic Editor: Xiaomei Qi Copyright 2013 Togguag Ni et al. This is a ope access article distributed uder the Creative Commos Attributio Licese, which permits urestricted use, distributio, ad reproductio i ay medium, provided the origial work is properly cited. Distributed deial of service (DDoS) attacks are oe of the major threats to the curret Iteret, ad applicatio-layer DDoS attacks utilizig legitimate HTTP requests to overwhelm victim resources are more udetectable. Cosequetly, either itrusio detectio systems (IDS) or victim server ca detect malicious packets. I this paper, a ovel approach to detect applicatio-layer DDoS attack is proposed based o etropy of HTTP GET requests per source IP address (HRPI). By approximatig the adaptive autoregressive (AAR) model, the HRPI time series is trasformed ito a multidimesioal vector series. The, a traied support vector machie (SVM) classifier is applied to idetify the attacks. The experimets with several databases are performed ad results show that this approach ca detect applicatio-layer DDoS attacks effectively. 1. Itroductio DDoSattackshavecausedseveredamagetoserversad will cause eve greater itimidatio to the developmet of ew Iteret services. DDoS attacks are categorized ito two classes: etwork-layer DDoS attacks ad applicatiolayer DDoS attacks. I etwork-layer DDoS attacks, attackers sed a large umber of bogus packets towards the victim server ad ormally attackers use IP spoofig. The victim server or IDS ca easily distiguish legitimate packets from DDoS packets. I cotrast, i applicatio-layer DDoS attacks, attackers attack the victim server through a flood of legitimate requests. I this attack model, attackers attack the victim Web servers by HTTP GET requests ad pullig large files from thevictimserverioverwhelmigumbers.also,attackers ca ru a massive umber of queries through the victim s search egie or database query to brig the server dow. To circumvet detectio, the attackers icreasigly move away from pure badwidth floods to stealthy DDoS attacks that masquerade as flash crowd. Flash crowd 1, 2] refers to the situatiowheaverylargeumberofuserssimultaeously access a website, which may be due to the aoucemet of a ew service or free software dowload. Because burst traffic ad high volume are the commo characteristics of applicatio-layer DDoS attacks ad flash crowd, it is ot easy to distiguish them. Therefore, applicatio layer DDoS attacks may be stealthier ad more dagerous for the websites tha the geeral etwork-layer DDoS attacks. Most well-kow DDoS coutermeasure 3] techiques are agaist etwork-layer DDoS attacks. Those techiques caot hadle applicatio-layer DDoS attacks. Couterig applicatio-layer DDoS attacks becomes a great challege. Statistical methods is used to detect characteristics of HTTP sessios ad employed rate-limitig as the primary defese mechaism i 4]. Costrait radom request attacks by the statistical methods are used to defed agaist the applicatiolayer DDoS attacks i 5]. A CAPTCHA puzzle is used to esure that the respose is geerated by a huma ot by a machie i 6]. A semi-markov model is proposed to describe the browsig Behaviors of Web surfers i 7], ad a improved semi-markov model is proposed to describe the dyamic behavior process of aggregated traffic i 8]. Recetly, trust-based methods 9, 10] wereitroducedfor resistig applicatio-layer DDoS attacks. The commo feature of these methods is that a defese system establishes credit records for each user. The credit value give to a seder isdesigedtobemeasuredbasedoitshistoryofcommuicatio patters. I applicatio-layer DDoS attacks, attack sources have bee programmed ad worked accordig to their attack
2 2 Cotrol Sciece ad Egieerig fuctios, so detectio based o its patter is possible. I this paper, the etropy of HTTP GET requests per source IP address (HRPI) is proposed, which reflects the essetial features of applicatio-layer DDoS attacks: the distributio of source IP address ad HTTP GET request frequecy. To icrease the detectio accuracy i various coditios, HRPI time series are trasformed ito a multidimesioal vector by estimatig the adaptive autoregressive (AAR) model parameters usig Kalma filter. Furthermore, a support vector machie (SVM) classifier, which is traied by AAR parameters of HRPI time series, is applied to classify the state of curret etwork traffic ad idetify the applicatio-layer DDoS attacks. The rest of the paper is orgaized as follows. Sectio 2 discusses the applicatio-layer DDoS attacks ad details their properties. Sectio 3 describes our approach to detect the applicatio-layer DDoSattacks.I Sectio 4, experimets are preseted to validate our detectio model. Fially, the coclusio is give i Sectio 5 ad it poits out the future work. 2. Applicatio-Layer DDoS Attacks Applicatio-layer DDoS attacks ca be clustered ito two types: badwidth exhaustig (HTTP floodig) ad resources exhaustig 11]. I badwidth exhaustig DDoS attacks, attackers attack the victim server through a flood of legitimate requests. Ay zombie machie has to establish a TCP coectio with the victim server, which requires a geuie IP address. Attacks maily focus o the homepage or a hot webpage, ad also differet web pages. I this case, the sources of the traffic coverge to a group of poits ad high HTTP Get request rate from the attackers. Besides the floodig attack patter, applicatio-layer DDoS attacks may focus o exhaustig the server resources such as Sockets, CPU, memory, disk/database badwidth, ad I/O badwidth. With icreasig computatioal complexity i Iteret applicatios ad larger etwork badwidth, server resources may become the bottleeck of these applicatios. Thistypeofattackisabletousefewerzombiesbuttheattack hasaevelargerdamagetothewebsite.however,thetraffic will be similar to the badwidth exhaustig DDoS. As a result, the sources of the traffic coverge to a group of poits but the targets of the traffic become dispersed i some extet. At thesametime,thefrequecyofhttpgetrequestfromthe attackersishighlylarge. O the Web, flash crowd refers to the situatio whe a very large umber of users simultaeously access a popular website, which produces a surge i traffic to the website ad mightcausethesitetobevirtuallyureachable.ddosattacks are absolutely differet from flash crowd, DDoS attacks are due to a icrease i the request rates for a small group of cliets while flash crowd is due to a icrease i the umber of cliets. The sources of flash crowd are defiitely scattered, coversely, the sources of applicatio-layer DDoS attacks coverge to a group of poits. 3. Our Approach 3.1. Defiitio of HRPI. For popular websites, the traffic targeted is a stream of successive HTTP Get requests. Defiitio 1. HTTPGetrequestsithecertaitimeiterval Δt isgiveitheformof (x 1,s 1 ), (x 2,s 2 ),...,(x,s ). For the (x i,s i ), x i is the source IP address ad s i is the umber of HTTP Get requests for x i. Defiitio 2. EtropyofHTTPGETrequestspersourceIP address (HRPI) is defied as SRE = p(x i ) lbp (x i ), (1) where p(x i ) is the probability of HTTP Get requests belogig to x i,adp(x i )=s i / i=1 s i. HRPI as a summarizatio tool is used to quatify the degree of dispersal or cocetratio of HTTP Get request feature distributios. Accordig to the aalysis i Sectio 2, we deduced the followig coclusio (DDoS as 1, ormal as 2, flash crowd as 3): HRPI (3) > HRPI (2) > HRPI (1). (2) I most cases, distributio form of source IP address of legitimateusersismoreuiformlyscatteredacrosstheiteret; the distributio form of source IP address of attackers is more cumulative i someplaces. I DDoS attacks, several clusters of source IP addresses ad lager umber of HTTP GET requests are coverged, so HRPI value dramatically drops whe attacks happe. Coversely, the sources of flash crowdarescatteredadtherewereosuchclusters,soitwill result i a abormal icrease i HRPI of the etwork Geeratio of HRPI Time Series. Adaptive autoregressive AAR (p)model12] of degree p is defied as y t = p k=1 a k t y t k +e t, (3) where y t deotes the observatio at istat t ad a k t deotes the time-varyig model parameters. As the traffic collectig device may cause measuremet errors, stochastic variable e t isusedtocapturethiserror.themodelusesaweightedsumof p previous values to estimate the curret observatio value. The weights a k t (k = 1,...,p) are time depedet, ad the curret value ca be predicted as a liear combiatio of p past values. By usig time-varyig AAR model, we allow a model of ormal behavior to adapt to the chages of the moitored system. Kalma filter is a adaptive ad recursive data processig algorithm that is suited for olie estimatio 13, 14]. Kalma filter ca process traffic matrix as a whole ad all traffic ca be estimated simultaeously. This implies that we do ot have to cosider all the previous data agai, to compute the optimal estimates; we oly eed to cosider the estimates from the previous time step ad the ew measuremet. I our case, we estimate the AAR model parameters from the observed alert series {y t }. The true parameters caot be observed directly ad i state space termiology they are called the state X. Now,assumethatwehaveaobservatio model givig the relatio betwee the uobservable state
3 Cotrol Sciece ad Egieerig 3 ad the observatios, ad a evolutio model describig the time-varyig ature of the state. So the AAR model ca be put i vector form as follows: Y t =H t X t +e t, (4) where H t deotes a iteral matrix ad H t =(Y t,...,y t p ). X t deotes the state vector at istat t ad X t =(a 1 t,...,ap t )T. Without prior iformatio, the evolutio of the state is ofte described with a radom walk model 15]. A liear equatio is costructed as follows to build a predictio model to correlate X t+1 ad X t : X t+1 =X t +w t, (5) where state oise w t ad measuremet oise e t are ucorrelated, zero-mea white-oise processes ad with covariace matrices σ 2 w ad σ2 e,respectively. ForrepresetigtheKalmafilterequatios, X deotes theestimateofx ad P t t 1 deotes error covariace matrix for estimatio error of the state at istat t usig observatios accumulated at istat t 1.Wheiitialcoditios, X 0 = EX 0 ] ad error covariace matrix P 0 = E( X 0 X 0 )( X 0 X 0 ) T ], the system state X t t 1 cabeestimated iterativelybythefollowigequatios: X t t 1 =X t 1, P t t 1 =P t 1 +C wt 1, K t =P t t 1 H T t (H tp t t 1 H T t +C et ) 1, X t = X t t 1 +K t (Y t H t X t t 1 ), P t =(I K t H t )P t t 1. Usig Kalma filter i practice requires iitial values for state X 0, error covariace P 0, state oise covariace C w (C w = σ 2 w I), ad observatio oise covariace C e (C e = σ 2 e = I). AcommoapproachistosetX 0 = 0, P 0 = I,adru the algorithm o a short segmet from observatio data backwards. The values obtaied i this way for X ad P 0 are the used to iitialize these values i the actual processig ru. The adaptatio speed icreases with C w ad the variace of state estimates is iversely proportioal to the value of C w. Therefore, it should be chose for a desired balace betwee state estimate variace ad filter adaptatio speed accordig to the applicatio. C w eeds to be set dyamically adaccordigiapplicatio Kalma Filter Smoothig. There are three classical smoothig algorithms, fixed-poit smoother, fixed-iterval smoother, ad fixed-lag smoother. We use fixed-lag smoother, sice it is suitable for olie processig whe a small, fixed delay of L observatios is allowed 16]. To estimate the state X t at istat t with a fixed-lag smoother,wewillwaittohaveobservatiosuptoistatt+l, where L>0.Thestateadobservatioequatioshaveow (6) exteded variables. The Kalma filter equatios remai the same, ad the observatio (4)becomes Y t =H t,0,...,0] X t X t 1. X t L The simplified state (5)ca be writte as X t+1 X t X t 1. X t (L 1) = ] ] X t X t 1 X t 2. X t L +e ] t. (7) ] w t (8) ]. ] ] 0 ] 3.4. SVM Classifier. By samplig the etwork traffic with time iterval Δt, calculatigthehrpiofeverysample,the HRPI sample series {HRPI i,i = 1,2,...,N} is gotte, N is the legth of the series. Based o (6) (8), multidimesioal vector X(i) of degree p cabeusedtodescribethestate features of etwork traffic. As a result, detectig DDoS attacks equates to classifyig X(i) series virtually. Support vector machie (SVM) is applied here, which is a well-kow data classificatio techique, to classify AAR parameters vector. SVM method ca get the optimal solutio whether the sample size teds to be fiite or ifiite. It ca establish a mappig of a oliear kerel fuctio, structurig the optimal hyperplae, so problem ca be coverted ito a liearly separable oe i the high-dimesioal feature space. Besides, it solves the dimesio problem ad its complexity has othig to do with the sample s dimesio. Sice traffic is oly cosidered as legitimate or attack, it is aturally a biary classificatio problem. The SVM classifier ca be described as η= M i=1 α i y i K(φ i,φ)+b, (9) where η is the classificatio result for the sample, α i is the Lagrage multiplies, y i is the category, ad y i { 1,1}. K(φ i,φ)is the kerel fuctio ad b is the deviatio factor. The optimal hyperplae that SVM classifier created i the high-dimesioal feature space is where f(φ)=sg ( α i y i (K (φ r,φ i )+K(φ s,φ i ))), (10) i SV b= 1 2 α i y i (K (φ r,φ i )+K(φ s,φ i )). (11) i SV SV (Support Vector) deotes the support vector ad φ r meas positive support vector, φ s meas egative support vector.
4 4 Cotrol Sciece ad Egieerig The coefficiet ca be obtaied by the followig quadratic programmig: max w (a) = i i=1a 1 a 2 i a j y i y j K(x i,x j ) i=1j=1 s.t. i=1 a i y i =0 0 α i C (i=1,2,...,m), (12) where C is the parameter to price the misclassificatio. Before the SVM ca classify traffics, it should udergo a traiig process to develop a classificatio model. We use the LibSVM library 17]to implemet SVM. 4. Experimets I order to evaluate the performace of our scheme, we divided our study ito two groups of experimets: to detect applicatio layer DDoS i ormal traffic ad i flash crowd Dataset. Normal traffic is the real-life Iteret traces collected from the traffic archive of Chagzhou uiversity WWW server. The traces cotai two weeks worth of all HTTP requests to the web server. We implemeted applicatio-layer DDoS attack i a simulator. Simulatios are carried out usig NS-2 etwork simulator o Liux platform. For geeratig attack traffic, there are 50 zombie machies ad a web server. Attack rates are 20 HTTP Get requests/s, 30 HTTP Get requests/s,...,60http Get requests/s, which simulate the attack rates of worm Mydoom, ad every attack lasts 1800s. Flash crowd is collected from the World Cup 98 website 18]. As this is a high arrival rate, we expect our approach to detect this traffic as flash crowd. We obtaied HRPI time series by multiple samplig ad calculatio whe the samplig iterval Δt is 0.1 s. As show i Figure 1(a), HRPI of ormal traffic varies with the time ad its mathematical expectatio is Figure 1(b) shows HRPI of DDoS attack ad its mathematical expectatio is 3.58, ad HRPI of flash crowd is show i Figure 1(c) with mathematical expectatio We ca see that the HRPI time series are sesitive to DDoS attack ad flash crowd, so HRPI ca distiguish three types of traffic distictly Model Parameters. There are three parameters which may affect the HRPI time series performaces. The first oe is the parameter p of AAR model. I practice, the model degree is ofte fixed usig some prior kowledge or guidelies. To optimize the goodess of fit verse, model complexity ratio, adalsotoeasethecomputatioalload,wesettledp=3as a degree which allowed the model to capture sufficietly well the ormal traffic behavior. The secod oe is state oise covariace C w (C w =σ 2 w I). The adaptatio speed of the Kalma filter is determied by the state oise covariace factor σ 2 w.itcotrolshowfast the state adopts the chages i observatios ad gives a suitable balace i adaptig to ormal behavior ad avoidig icorporatig aomalous behavior i to the model. We HRPI HRPI HRPI (s) (a) HRPIoformaltraffic (s) (b) HRPI of DDoS attack (s) (c) HRPI of flash crowd Figure 1: HRPI of differet web traffics. experimeted with differet values ad chose to use σ 2 w = The third oe is the lag of Kalma filter, ad we oticed a sigificat icrease i model accuracy whe L = 1.The icrease i accuracy was slower whe further icreasig L. As larger L meas also loger delay i detectio, we chose to use L= Experimets ad Results Evaluatio Criteria. I this paper, a group of performace metrics i classificatio problems are used for the evaluatio of the results, cosistig of FPR, FNR, accuracy,
5 Cotrol Sciece ad Egieerig 5 Table 1: The results of DDoS detectio i ormal traffic. Accuracy FPR FNR Precisio Recall ROC P % 8.24% 7.27% 91.62% 92.51% 90.12% P % 5.05% 3.94% 94.51% 95.28% 95.31% P % 3.74% 3.08% 96.79% 97.00% 98.79% P % 2.76% 2.23% 97.30% 97.64% 99.10% P % 2.05% 1.88% 97.94% 98.02% 99.26% Table 2: The results of DDoS detectio i flash crowd. Accuracy FPR FNR Precisio Recall ROC T % 6.69% 6.03% 92.66% 93.28% 91.67% T % 4.10% 3.09% 95.64% 96.37% 96.34% T % 3.52% 2.58% 96.82% 96.95% 98.99% T % 2.16% 1.94% 97.22% 98.03% 99.61% T % 1.47% 1.02% 98.29% 98.49% 99.70% precisio, recall, ad ROC. Let TP represet the ormal test samples that have bee correctly classified ad let FP represet the oes that have bee wrogly classified. Let TN represet the attackig test samples that have bee correctly classified ad let FN represet the oes that have bee falsely classified. Thus, the False-Positive Rate (FPR) ad the False- Negative Rate (FNR) are the proportios of wrogly classified ormal test samples ad attackig test samples, respectively (FPR = FP/(FP + TN), FNR = FN/(TP + FN)). Accuracy states the overall percetage of correctly classified attackig test samples (accuracy = (TP + TN)/(TP + FP + TN + FN)). Precisio as the classifier s safety, states the degree i which messages idetified as attackig test samples are ideed malicious (precisio = TP/(TP + FP)). Recall as the classifier s effectiveess, states the percetage of attackig test samples that the classifier maages to classify correctly (recall = TP/(TP + FN)). Receiver Operatig Characteristic (ROC) as a classifier s balace ability betwee its FPR ad its FNR is a fuctio of varyig classificatio threshold Experimet 1: Detect DDoS Attacks i Normal Traffic. We set that the samplig iterval Δt is 0.1 s, HRPI time series legth N is 100, so the detectio time is 10 s. I this experimet, ormal traffic cotai 600 series, ad DDoS attack traffic cotai 450 series. Obtaied dataset is divided ito two parts: traiig data cotais 60% of total data values, testig data cotais the rest of the obtaied dataset. The kerel fuctio i SVM classifier is radial basis fuctio (RBF) ad the robustess of the classifiers is evaluated usig 10-fold cross-validatio. I order to test the robustess of our method to the disturbace of ormal traffic, we do five experimets amed as P-20, P-30,..., P-60, i which traffic attacks are 20 HTTP Get requests/s,...,60httpget requests/s mixig ormal traffic at the same time. Table 1 shows the performace results; the detectio ratio of our approach icreases whe the attack traffic volume icreases. Whe ormal traffic is much larger tha attack traffic, the detectio ratio still keeps a high level. This meas that our approach ca idetify the DDoS attack traffic with a high precisio, ad be sesitive to DDoS attack traffic Experimet 2: Detect DDoS Attacks i Flash Crowd. I this experimet, the samplig iterval Δt is 0.1 s, ad HRPI time series legth N is 100, too. We sample flash crowd 500 series, ad DDoS attack traffic 350 series. The traiig ad testig method of SVM is the same as experimet 1. We do five experimets amed as T-20, T-30,...,T-60, by mixig attackig traffic 20 HTTP Get requests/s,...,60 HTTP Get requests/s ad flash crowd. Table 2 shows the performace results, with the icremet of flash crowd, the detectio ratio of our approach does ot declie rapidly. The FPR ad FNR are reduced with the icrease of attack rate, ad the accuracy, precisio, recall, ad ROC are asceded with the icrease of attack rate. I the above two groups of experimets, the false egativescom yfromtwoaspects:firstly,duetotheicrease of ormal traffic or flash crowd, which makes the HRPI states lear to ormal oes, thus makig the differece too small for detectio. Secodly, the etwork state shift caused by etwork radom oise results i false egative. 5. Coclusio Applicatio-layer DDoS attacks detectio is a hot ad difficult research topic i the field of itrusio detectio. Based o the characteristics of DDoS attack, this paper proposes a ovel approach to detect DDoS attacks. The work provides two cotributios: (1) HRPI is itroduced to detect DDoS attacks, ad it reflects the essetial features of attacks ad (2) a detectio scheme agaist DDoS attacks is proposed, ad it ca achieve high detectio efficiecy ad flexibility. I our future work, we will make a detailed study of how to set all kids of parameters i differet applicatio scearios adaptively. Ackowledgmet This work was supported by the Natioal Natural Sciece Foudatio of Chia uder Cotact ( ). Refereces 1] T. Thapgam, S. Yu, W. Zhou, ad G. Beliakov, Discrimiatig DDoS attack traffic from flash crowd through packet arrival patters, i Proceedigs of the IEEE Coferece o Computer Commuicatios Workshops (INFOCOM 11), pp , April ] G. Oikoomou ad J. Mirkovic, Modelig huma behavior for defese agaist flash-crowd attacks, i Proceedigs of the IEEE Iteratioal Coferece o Commuicatios (ICC 09),pp.1 6, Jue ] H. Beitollahi ad G. Decoick, Aalyzig well-kow coutermeasures agaist distributed deial of service attacks, Computer Commuicatios, vol. 35, pp , ] S. Raja, R. Swamiatha, M. Uysal, ad E. Kightly, DDoSresiliet schedulig to couter applicatio layer attacks uder
6 6 Cotrol Sciece ad Egieerig imperfect detectio, i Proceedigs of the 25th IEEE Iteratioal Coferece o Computer Commuicatios (INFOCOM 06),pp.1 13,April ] W. Ye ad M.-F. Lee, Defedig applicatio DDoS with costrait radom request attacks, i Proceedigs of the Asia- Pacific Coferece o Commuicatios, pp , Perth, Australia, October ] L. Vo Ah, M. Blum, ad J. Lagford, Tellig humas ad computers apart automatically, Commuicatios of the ACM, vol.47,o.2,pp.56 60, ] Y. Xie ad S.-Z. Yu, A large-scale hidde semi-markov model for aomaly detectio o user browsig behaviors, IEEE/ACM Trasactios o Networkig,vol.17,o.1,pp.54 65, ] Y.Xie,S.Tag,adX.Huag, Detectiglatetattackbehavior from aggregated Web traffic, Computer Commuicatios,o.5, pp , ] J.Yu,C.Fag,L.Luetal., Alightweightmechaismtomitigate applicatio layer DDoS attacks, Scalable Iformatio Systems, vol. 18, pp , ] P. Du ad A. Nakao, OverCourt: DDoS mitigatio through credit-based traffic segregatio ad path migratio, Computer Commuicatios,vol.33,o.18,pp , ] H. Beitollahi ad G. Decoick, Tacklig Applicatio-layer DDoS Attacks, Procedia Computer Sciece,vol.10,pp , ] Q.-D. Su, D.-Y. Zhag, ad P. Gao, Detectig distributed deial of service attacks based o time series aalysis, Chiese Computers,vol.28,o.5,pp , ] R. Ya, Q. Zheg, ad H. Li, Combiig adaptive filterig adifflowstodetectddosattackswithiarouter, KSII Trasactios o Iteret ad Iformatio Systems, vol.4,o.3, pp , ] S. We, W. Jia, W. Zhou, W. Zhou, ad C. Xu, CALD: Survivig various applicatio-layer DDoS attacks that mimic flash crowd, i Proceedigs of the 4th Iteratioal Coferece o Network ad System Security (NSS 10), pp , Victoria, Australia, September ] S. Haykl, Adaptive Filter Theory, Pretice-Hall, Upper saddle River, NJ, USA, 3rd editio, ] J. Viiikka, H. Debar, L. Mé, A. Lehikoie, ad M. Tarvaie, Processig itrusio detectio alert aggregates with time series modelig, Iformatio Fusio,vol.10,o.4,pp , ] J. Platt, Sequetial miimal optimizatio: a fast algorithm for traiig support vector machies, Tech. Rep. MSR-TR-98-14, Microsoft Research, ] M. Arlitt ad T. Ji, 1998 World Cup Web Site Access Logs, 1998,
7 Rotatig Machiery Egieerig The Scietific World Joural Distributed Sesor Networks Sesors Cotrol Sciece ad Egieerig Advaces i Civil Egieerig Submit your mauscripts at Electrical ad Computer Egieerig Robotics VLSI Desig Advaces i OptoElectroics Navigatio ad Observatio Chemical Egieerig Active ad Passive Electroic Compoets Ateas ad Propagatio Aerospace Egieerig Modellig & Simulatio i Egieerig Shock ad Vibratio Advaces i Acoustics ad Vibratio
Modified Line Search Method for Global Optimization
Modified Lie Search Method for Global Optimizatio Cria Grosa ad Ajith Abraham Ceter of Excellece for Quatifiable Quality of Service Norwegia Uiversity of Sciece ad Techology Trodheim, Norway {cria, ajith}@q2s.tu.o
Research Article Sign Data Derivative Recovery
Iteratioal Scholarly Research Network ISRN Applied Mathematics Volume 0, Article ID 63070, 7 pages doi:0.540/0/63070 Research Article Sig Data Derivative Recovery L. M. Housto, G. A. Glass, ad A. D. Dymikov
Study on the application of the software phase-locked loop in tracking and filtering of pulse signal
Advaced Sciece ad Techology Letters, pp.31-35 http://dx.doi.org/10.14257/astl.2014.78.06 Study o the applicatio of the software phase-locked loop i trackig ad filterig of pulse sigal Sog Wei Xia 1 (College
Output Analysis (2, Chapters 10 &11 Law)
B. Maddah ENMG 6 Simulatio 05/0/07 Output Aalysis (, Chapters 10 &11 Law) Comparig alterative system cofiguratio Sice the output of a simulatio is radom, the comparig differet systems via simulatio should
LECTURE 13: Cross-validation
LECTURE 3: Cross-validatio Resampli methods Cross Validatio Bootstrap Bias ad variace estimatio with the Bootstrap Three-way data partitioi Itroductio to Patter Aalysis Ricardo Gutierrez-Osua Texas A&M
An Adaptive Method for Source-end Detection of Pulsing DoS Attacks
, pp.279-288 http://dx.doi.org/10.14257/ijsia.2013.7.5.26 A Adaptive Method for Source-ed Detectio of Pulsig DoS Attacks Mig Yu School of Iformatio ad Commuicatio Egieerig, Dalia Uiversity of Techology,
Confidence Intervals for One Mean
Chapter 420 Cofidece Itervals for Oe Mea Itroductio This routie calculates the sample size ecessary to achieve a specified distace from the mea to the cofidece limit(s) at a stated cofidece level for a
Project Deliverables. CS 361, Lecture 28. Outline. Project Deliverables. Administrative. Project Comments
Project Deliverables CS 361, Lecture 28 Jared Saia Uiversity of New Mexico Each Group should tur i oe group project cosistig of: About 6-12 pages of text (ca be loger with appedix) 6-12 figures (please
Properties of MLE: consistency, asymptotic normality. Fisher information.
Lecture 3 Properties of MLE: cosistecy, asymptotic ormality. Fisher iformatio. I this sectio we will try to uderstad why MLEs are good. Let us recall two facts from probability that we be used ofte throughout
Non-life insurance mathematics. Nils F. Haavardsson, University of Oslo and DNB Skadeforsikring
No-life isurace mathematics Nils F. Haavardsso, Uiversity of Oslo ad DNB Skadeforsikrig Mai issues so far Why does isurace work? How is risk premium defied ad why is it importat? How ca claim frequecy
Data Analysis and Statistical Behaviors of Stock Market Fluctuations
44 JOURNAL OF COMPUTERS, VOL. 3, NO. 0, OCTOBER 2008 Data Aalysis ad Statistical Behaviors of Stock Market Fluctuatios Ju Wag Departmet of Mathematics, Beijig Jiaotog Uiversity, Beijig 00044, Chia Email:
(VCP-310) 1-800-418-6789
Maual VMware Lesso 1: Uderstadig the VMware Product Lie I this lesso, you will first lear what virtualizatio is. Next, you ll explore the products offered by VMware that provide virtualizatio services.
Vladimir N. Burkov, Dmitri A. Novikov MODELS AND METHODS OF MULTIPROJECTS MANAGEMENT
Keywords: project maagemet, resource allocatio, etwork plaig Vladimir N Burkov, Dmitri A Novikov MODELS AND METHODS OF MULTIPROJECTS MANAGEMENT The paper deals with the problems of resource allocatio betwee
5: Introduction to Estimation
5: Itroductio to Estimatio Cotets Acroyms ad symbols... 1 Statistical iferece... Estimatig µ with cofidece... 3 Samplig distributio of the mea... 3 Cofidece Iterval for μ whe σ is kow before had... 4 Sample
Review: Classification Outline
Data Miig CS 341, Sprig 2007 Decisio Trees Neural etworks Review: Lecture 6: Classificatio issues, regressio, bayesia classificatio Pretice Hall 2 Data Miig Core Techiques Classificatio Clusterig Associatio
Chapter 6: Variance, the law of large numbers and the Monte-Carlo method
Chapter 6: Variace, the law of large umbers ad the Mote-Carlo method Expected value, variace, ad Chebyshev iequality. If X is a radom variable recall that the expected value of X, E[X] is the average value
VEHICLE TRACKING USING KALMAN FILTER AND FEATURES
Sigal & Image Processig : A Iteratioal Joural (SIPIJ) Vol.2, No.2, Jue 2011 VEHICLE TRACKING USING KALMAN FILTER AND FEATURES Amir Salarpour 1 ad Arezoo Salarpour 2 ad Mahmoud Fathi 2 ad MirHossei Dezfoulia
Soving Recurrence Relations
Sovig Recurrece Relatios Part 1. Homogeeous liear 2d degree relatios with costat coefficiets. Cosider the recurrece relatio ( ) T () + at ( 1) + bt ( 2) = 0 This is called a homogeeous liear 2d degree
Research Article An Approach to Evaluating Computer Network Security with Intuitionistic Trapezoidal Fuzzy Information
Joural of Cotrol Sciece ad Egieerig, Article ID 604920, 4 pages http://dx.doi.org/10.1155/2014/604920 Research Article A Approach to Evaluatig Computer Network Security with Ituitioistic Trapezoidal Fuzzy
Systems Design Project: Indoor Location of Wireless Devices
Systems Desig Project: Idoor Locatio of Wireless Devices Prepared By: Bria Murphy Seior Systems Sciece ad Egieerig Washigto Uiversity i St. Louis Phoe: (805) 698-5295 Email: [email protected] Supervised
Chapter 7 - Sampling Distributions. 1 Introduction. What is statistics? It consist of three major areas:
Chapter 7 - Samplig Distributios 1 Itroductio What is statistics? It cosist of three major areas: Data Collectio: samplig plas ad experimetal desigs Descriptive Statistics: umerical ad graphical summaries
Incremental calculation of weighted mean and variance
Icremetal calculatio of weighted mea ad variace Toy Fich [email protected] [email protected] Uiversity of Cambridge Computig Service February 009 Abstract I these otes I eplai how to derive formulae for umerically
A probabilistic proof of a binomial identity
A probabilistic proof of a biomial idetity Joatho Peterso Abstract We give a elemetary probabilistic proof of a biomial idetity. The proof is obtaied by computig the probability of a certai evet i two
PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM
PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY Physical ad Mathematical Scieces 2015, 1, p. 15 19 M a t h e m a t i c s AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM A. G. GULYAN Chair of Actuarial Mathematics
Overview of some probability distributions.
Lecture Overview of some probability distributios. I this lecture we will review several commo distributios that will be used ofte throughtout the class. Each distributio is usually described by its probability
Reliability Analysis in HPC clusters
Reliability Aalysis i HPC clusters Narasimha Raju, Gottumukkala, Yuda Liu, Chokchai Box Leagsuksu 1, Raja Nassar, Stephe Scott 2 College of Egieerig & Sciece, Louisiaa ech Uiversity Oak Ridge Natioal Lab
Chapter 7 Methods of Finding Estimators
Chapter 7 for BST 695: Special Topics i Statistical Theory. Kui Zhag, 011 Chapter 7 Methods of Fidig Estimators Sectio 7.1 Itroductio Defiitio 7.1.1 A poit estimator is ay fuctio W( X) W( X1, X,, X ) of
A gentle introduction to Expectation Maximization
A getle itroductio to Expectatio Maximizatio Mark Johso Brow Uiversity November 2009 1 / 15 Outlie What is Expectatio Maximizatio? Mixture models ad clusterig EM for setece topic modelig 2 / 15 Why Expectatio
PSYCHOLOGICAL STATISTICS
UNIVERSITY OF CALICUT SCHOOL OF DISTANCE EDUCATION B Sc. Cousellig Psychology (0 Adm.) IV SEMESTER COMPLEMENTARY COURSE PSYCHOLOGICAL STATISTICS QUESTION BANK. Iferetial statistics is the brach of statistics
Plug-in martingales for testing exchangeability on-line
Plug-i martigales for testig exchageability o-lie Valetia Fedorova, Alex Gammerma, Ilia Nouretdiov, ad Vladimir Vovk Computer Learig Research Cetre Royal Holloway, Uiversity of Lodo, UK {valetia,ilia,alex,vovk}@cs.rhul.ac.uk
Normal Distribution.
Normal Distributio www.icrf.l Normal distributio I probability theory, the ormal or Gaussia distributio, is a cotiuous probability distributio that is ofte used as a first approimatio to describe realvalued
Traffic Modeling and Prediction using ARIMA/GARCH model
Traffic Modelig ad Predictio usig ARIMA/GARCH model Preseted by Zhili Su, Bo Zhou, Uiversity of Surrey, UK COST 285 Symposium 8-9 September 2005 Muich, Germay Outlie Motivatio ARIMA/GARCH model Parameter
Analyzing Longitudinal Data from Complex Surveys Using SUDAAN
Aalyzig Logitudial Data from Complex Surveys Usig SUDAAN Darryl Creel Statistics ad Epidemiology, RTI Iteratioal, 312 Trotter Farm Drive, Rockville, MD, 20850 Abstract SUDAAN: Software for the Statistical
Estimating Probability Distributions by Observing Betting Practices
5th Iteratioal Symposium o Imprecise Probability: Theories ad Applicatios, Prague, Czech Republic, 007 Estimatig Probability Distributios by Observig Bettig Practices Dr C Lych Natioal Uiversity of Irelad,
COMPARISON OF THE EFFICIENCY OF S-CONTROL CHART AND EWMA-S 2 CONTROL CHART FOR THE CHANGES IN A PROCESS
COMPARISON OF THE EFFICIENCY OF S-CONTROL CHART AND EWMA-S CONTROL CHART FOR THE CHANGES IN A PROCESS Supraee Lisawadi Departmet of Mathematics ad Statistics, Faculty of Sciece ad Techoology, Thammasat
Multi-server Optimal Bandwidth Monitoring for QoS based Multimedia Delivery Anup Basu, Irene Cheng and Yinzhe Yu
Multi-server Optimal Badwidth Moitorig for QoS based Multimedia Delivery Aup Basu, Iree Cheg ad Yizhe Yu Departmet of Computig Sciece U. of Alberta Architecture Applicatio Layer Request receptio -coectio
Recovery time guaranteed heuristic routing for improving computation complexity in survivable WDM networks
Computer Commuicatios 30 (2007) 1331 1336 wwwelseviercom/locate/comcom Recovery time guarateed heuristic routig for improvig computatio complexity i survivable WDM etworks Lei Guo * College of Iformatio
DDoS attacks defence strategies based on nonparametric CUSUM algorithm
Abstract DDoS attacks defece strategies based o oparametric CUSUM algorithm Chaghog Ya 1*, Qi Dog 2, Hog Wag 3 1 School of Iformatio Egieerig, Yacheg Istitute of Techology, No.9 XiWag Aveue Road, Yacheg,
ADAPTIVE NETWORKS SAFETY CONTROL ON FUZZY LOGIC
8 th Iteratioal Coferece o DEVELOPMENT AND APPLICATION SYSTEMS S u c e a v a, R o m a i a, M a y 25 27, 2 6 ADAPTIVE NETWORKS SAFETY CONTROL ON FUZZY LOGIC Vadim MUKHIN 1, Elea PAVLENKO 2 Natioal Techical
Evaluating Model for B2C E- commerce Enterprise Development Based on DEA
, pp.180-184 http://dx.doi.org/10.14257/astl.2014.53.39 Evaluatig Model for B2C E- commerce Eterprise Developmet Based o DEA Weli Geg, Jig Ta Computer ad iformatio egieerig Istitute, Harbi Uiversity of
0.7 0.6 0.2 0 0 96 96.5 97 97.5 98 98.5 99 99.5 100 100.5 96.5 97 97.5 98 98.5 99 99.5 100 100.5
Sectio 13 Kolmogorov-Smirov test. Suppose that we have a i.i.d. sample X 1,..., X with some ukow distributio P ad we would like to test the hypothesis that P is equal to a particular distributio P 0, i.e.
Basic Measurement Issues. Sampling Theory and Analog-to-Digital Conversion
Theory ad Aalog-to-Digital Coversio Itroductio/Defiitios Aalog-to-digital coversio Rate Frequecy Aalysis Basic Measuremet Issues Reliability the extet to which a measuremet procedure yields the same results
The analysis of the Cournot oligopoly model considering the subjective motive in the strategy selection
The aalysis of the Courot oligopoly model cosiderig the subjective motive i the strategy selectio Shigehito Furuyama Teruhisa Nakai Departmet of Systems Maagemet Egieerig Faculty of Egieerig Kasai Uiversity
Case Study. Normal and t Distributions. Density Plot. Normal Distributions
Case Study Normal ad t Distributios Bret Halo ad Bret Larget Departmet of Statistics Uiversity of Wiscosi Madiso October 11 13, 2011 Case Study Body temperature varies withi idividuals over time (it ca
SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES
SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES Read Sectio 1.5 (pages 5 9) Overview I Sectio 1.5 we lear to work with summatio otatio ad formulas. We will also itroduce a brief overview of sequeces,
Overview on S-Box Design Principles
Overview o S-Box Desig Priciples Debdeep Mukhopadhyay Assistat Professor Departmet of Computer Sciece ad Egieerig Idia Istitute of Techology Kharagpur INDIA -721302 What is a S-Box? S-Boxes are Boolea
Clustering Algorithm Analysis of Web Users with Dissimilarity and SOM Neural Networks
JONAL OF SOFTWARE, VOL. 7, NO., NOVEMBER 533 Clusterig Algorithm Aalysis of Web Users with Dissimilarity ad SOM Neal Networks Xiao Qiag School of Ecoomics ad maagemet, Lazhou Jiaotog Uiversity, Lazhou;
, a Wishart distribution with n -1 degrees of freedom and scale matrix.
UMEÅ UNIVERSITET Matematisk-statistiska istitutioe Multivariat dataaalys D MSTD79 PA TENTAMEN 004-0-9 LÖSNINGSFÖRSLAG TILL TENTAMEN I MATEMATISK STATISTIK Multivariat dataaalys D, 5 poäg.. Assume that
Taking DCOP to the Real World: Efficient Complete Solutions for Distributed Multi-Event Scheduling
Taig DCOP to the Real World: Efficiet Complete Solutios for Distributed Multi-Evet Schedulig Rajiv T. Maheswara, Milid Tambe, Emma Bowrig, Joatha P. Pearce, ad Pradeep araatham Uiversity of Souther Califoria
Spam Detection. A Bayesian approach to filtering spam
Spam Detectio A Bayesia approach to filterig spam Kual Mehrotra Shailedra Watave Abstract The ever icreasig meace of spam is brigig dow productivity. More tha 70% of the email messages are spam, ad it
1. C. The formula for the confidence interval for a population mean is: x t, which was
s 1. C. The formula for the cofidece iterval for a populatio mea is: x t, which was based o the sample Mea. So, x is guarateed to be i the iterval you form.. D. Use the rule : p-value
Hypothesis testing. Null and alternative hypotheses
Hypothesis testig Aother importat use of samplig distributios is to test hypotheses about populatio parameters, e.g. mea, proportio, regressio coefficiets, etc. For example, it is possible to stipulate
Hypergeometric Distributions
7.4 Hypergeometric Distributios Whe choosig the startig lie-up for a game, a coach obviously has to choose a differet player for each positio. Similarly, whe a uio elects delegates for a covetio or you
1 Computing the Standard Deviation of Sample Means
Computig the Stadard Deviatio of Sample Meas Quality cotrol charts are based o sample meas ot o idividual values withi a sample. A sample is a group of items, which are cosidered all together for our aalysis.
Maximum Likelihood Estimators.
Lecture 2 Maximum Likelihood Estimators. Matlab example. As a motivatio, let us look at oe Matlab example. Let us geerate a radom sample of size 00 from beta distributio Beta(5, 2). We will lear the defiitio
Evaluation of Different Fitness Functions for the Evolutionary Testing of an Autonomous Parking System
Evaluatio of Differet Fitess Fuctios for the Evolutioary Testig of a Autoomous Parkig System Joachim Wegeer 1, Oliver Bühler 2 1 DaimlerChrysler AG, Research ad Techology, Alt-Moabit 96 a, D-1559 Berli,
Iran. J. Chem. Chem. Eng. Vol. 26, No.1, 2007. Sensitivity Analysis of Water Flooding Optimization by Dynamic Optimization
Ira. J. Chem. Chem. Eg. Vol. 6, No., 007 Sesitivity Aalysis of Water Floodig Optimizatio by Dyamic Optimizatio Gharesheiklou, Ali Asghar* + ; Mousavi-Dehghai, Sayed Ali Research Istitute of Petroleum Idustry
Automatic Tuning for FOREX Trading System Using Fuzzy Time Series
utomatic Tuig for FOREX Tradig System Usig Fuzzy Time Series Kraimo Maeesilp ad Pitihate Soorasa bstract Efficiecy of the automatic currecy tradig system is time depedet due to usig fixed parameters which
Research Method (I) --Knowledge on Sampling (Simple Random Sampling)
Research Method (I) --Kowledge o Samplig (Simple Radom Samplig) 1. Itroductio to samplig 1.1 Defiitio of samplig Samplig ca be defied as selectig part of the elemets i a populatio. It results i the fact
Domain 1: Designing a SQL Server Instance and a Database Solution
Maual SQL Server 2008 Desig, Optimize ad Maitai (70-450) 1-800-418-6789 Domai 1: Desigig a SQL Server Istace ad a Database Solutio Desigig for CPU, Memory ad Storage Capacity Requiremets Whe desigig a
Your organization has a Class B IP address of 166.144.0.0 Before you implement subnetting, the Network ID and Host ID are divided as follows:
Subettig Subettig is used to subdivide a sigle class of etwork i to multiple smaller etworks. Example: Your orgaizatio has a Class B IP address of 166.144.0.0 Before you implemet subettig, the Network
Overview. Learning Objectives. Point Estimate. Estimation. Estimating the Value of a Parameter Using Confidence Intervals
Overview Estimatig the Value of a Parameter Usig Cofidece Itervals We apply the results about the sample mea the problem of estimatio Estimatio is the process of usig sample data estimate the value of
A Faster Clause-Shortening Algorithm for SAT with No Restriction on Clause Length
Joural o Satisfiability, Boolea Modelig ad Computatio 1 2005) 49-60 A Faster Clause-Shorteig Algorithm for SAT with No Restrictio o Clause Legth Evgey Datsi Alexader Wolpert Departmet of Computer Sciece
Effective Hybrid Intrusion Detection System: A Layered Approach
I. J. Computer Network ad Iformatio Security, 2015, 3, 35-41 Published Olie February 2015 i MECS (http://www.mecs-press.org/) DOI: 10.5815/ijcis.2015.03.05 Effective Hybrid Itrusio Detectio System: A Layered
Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations
Real-Time Computig Without Stable States: A New Framework for Neural Computatio Based o Perturbatios Wolfgag aass+, Thomas Natschläger+ & Hery arkram* + Istitute for Theoretical Computer Sciece, Techische
Z-TEST / Z-STATISTIC: used to test hypotheses about. µ when the population standard deviation is unknown
Z-TEST / Z-STATISTIC: used to test hypotheses about µ whe the populatio stadard deviatio is kow ad populatio distributio is ormal or sample size is large T-TEST / T-STATISTIC: used to test hypotheses about
I. Chi-squared Distributions
1 M 358K Supplemet to Chapter 23: CHI-SQUARED DISTRIBUTIONS, T-DISTRIBUTIONS, AND DEGREES OF FREEDOM To uderstad t-distributios, we first eed to look at aother family of distributios, the chi-squared distributios.
CHAPTER 3 DIGITAL CODING OF SIGNALS
CHAPTER 3 DIGITAL CODING OF SIGNALS Computers are ofte used to automate the recordig of measuremets. The trasducers ad sigal coditioig circuits produce a voltage sigal that is proportioal to a quatity
Optimal Adaptive Bandwidth Monitoring for QoS Based Retrieval
1 Optimal Adaptive Badwidth Moitorig for QoS Based Retrieval Yizhe Yu, Iree Cheg ad Aup Basu (Seior Member) Departmet of Computig Sciece Uiversity of Alberta Edmoto, AB, T6G E8, CAADA {yizhe, aup, li}@cs.ualberta.ca
Log-Logistic Software Reliability Growth Model
Log-Logistic Software Reliability Growth Model Swapa S. Gokhale ad Kishor S. Trivedi 2y Bours College of Egg. CACC, Dept. of ECE Uiversity of Califoria Duke Uiversity Riverside, CA 9252 Durham, NC 2778-29
Enhancing Oracle Business Intelligence with cubus EV How users of Oracle BI on Essbase cubes can benefit from cubus outperform EV Analytics (cubus EV)
Ehacig Oracle Busiess Itelligece with cubus EV How users of Oracle BI o Essbase cubes ca beefit from cubus outperform EV Aalytics (cubus EV) CONTENT 01 cubus EV as a ehacemet to Oracle BI o Essbase 02
Center, Spread, and Shape in Inference: Claims, Caveats, and Insights
Ceter, Spread, ad Shape i Iferece: Claims, Caveats, ad Isights Dr. Nacy Pfeig (Uiversity of Pittsburgh) AMATYC November 2008 Prelimiary Activities 1. I would like to produce a iterval estimate for the
Determining the sample size
Determiig the sample size Oe of the most commo questios ay statisticia gets asked is How large a sample size do I eed? Researchers are ofte surprised to fid out that the aswer depeds o a umber of factors
Cantilever Beam Experiment
Mechaical Egieerig Departmet Uiversity of Massachusetts Lowell Catilever Beam Experimet Backgroud A disk drive maufacturer is redesigig several disk drive armature mechaisms. This is the result of evaluatio
CONTROL CHART BASED ON A MULTIPLICATIVE-BINOMIAL DISTRIBUTION
www.arpapress.com/volumes/vol8issue2/ijrras_8_2_04.pdf CONTROL CHART BASED ON A MULTIPLICATIVE-BINOMIAL DISTRIBUTION Elsayed A. E. Habib Departmet of Statistics ad Mathematics, Faculty of Commerce, Beha
Nr. 2. Interpolation of Discount Factors. Heinz Cremers Willi Schwarz. Mai 1996
Nr 2 Iterpolatio of Discout Factors Heiz Cremers Willi Schwarz Mai 1996 Autore: Herausgeber: Prof Dr Heiz Cremers Quatitative Methode ud Spezielle Bakbetriebslehre Hochschule für Bakwirtschaft Dr Willi
Statistical and Fuzzy Approach for Database Security
tatistical ad Fuzzy Approach for Database ecurity Gag Lu, Jukai Yi chool of Iformatio ciece ad Techology Beijig iversity of Chemical Techology Beijig 100029, Chia [email protected] Kevi Lü Bruel iversity,
AN ANOMALY DETECTION SYSTEM FOR DDOS ATTACK IN GRID COMPUTING
AN ANOMALY DETECTION SYSTEM FOR DDOS ATTACK IN GRID COMPUTING 1 Sumit kar, 2 Bibhudatta sahoo 1,2 NIT Rourkela/CSE, Rourkela, Idia [email protected], [email protected] ABSTRACT: Grid computig is
MARKOV MODEL M/M/M/K IN CONTACT CENTER
MARKOV MODEL M/M/M/K IN CONTACT CENTER Erik CHROMY 1, Ja DIEZKA 1, Matej KAVACKY 1 1 Istitute of Telecommuicatios, Faculty of Electrical Egieerig ad Iformatio Techology, Slovak Uiversity of Techology Bratislava,
Verifying the Availability of Cloud Applications
Melaie Siebehaar, Olga Wege, Roy Has, Hasa Terca, Ralf Steimetz: Verifyig the Availability of Cloud Applicatios. I: Proceedigs of the 3rd Iteratioal Coferece o Cloud Computig ad Services Sciece (CLOSER
Extracting Similar and Opposite News Websites Based on Sentiment Analysis
202 Iteratioal Coferece o Idustrial ad Itelliget Iformatio (ICIII 202) IPCSIT vol.3 (202) (202) IACSIT Press, Sigapore Extractig Similar ad Opposite ews Websites Based o Setimet Aalysis Jiawei Zhag, Yukiko
How To Improve Software Reliability
2 Iteratioal Joural of Computer Applicatios (975 8887) A Software Reliability Growth Model for Three-Tier Cliet Server System Pradeep Kumar Iformatio Techology Departmet ABES Egieerig College, Ghaziabad
Escola Federal de Engenharia de Itajubá
Escola Federal de Egeharia de Itajubá Departameto de Egeharia Mecâica Pós-Graduação em Egeharia Mecâica MPF04 ANÁLISE DE SINAIS E AQUISÇÃO DE DADOS SINAIS E SISTEMAS Trabalho 02 (MATLAB) Prof. Dr. José
Confidence Intervals. CI for a population mean (σ is known and n > 30 or the variable is normally distributed in the.
Cofidece Itervals A cofidece iterval is a iterval whose purpose is to estimate a parameter (a umber that could, i theory, be calculated from the populatio, if measuremets were available for the whole populatio).
Modeling of Ship Propulsion Performance
odelig of Ship Propulsio Performace Bejami Pjedsted Pederse (FORCE Techology, Techical Uiversity of Demark) Ja Larse (Departmet of Iformatics ad athematical odelig, Techical Uiversity of Demark) Full scale
Annuities Under Random Rates of Interest II By Abraham Zaks. Technion I.I.T. Haifa ISRAEL and Haifa University Haifa ISRAEL.
Auities Uder Radom Rates of Iterest II By Abraham Zas Techio I.I.T. Haifa ISRAEL ad Haifa Uiversity Haifa ISRAEL Departmet of Mathematics, Techio - Israel Istitute of Techology, 3000, Haifa, Israel I memory
DAME - Microsoft Excel add-in for solving multicriteria decision problems with scenarios Radomir Perzina 1, Jaroslav Ramik 2
Itroductio DAME - Microsoft Excel add-i for solvig multicriteria decisio problems with scearios Radomir Perzia, Jaroslav Ramik 2 Abstract. The mai goal of every ecoomic aget is to make a good decisio,
A Combined Continuous/Binary Genetic Algorithm for Microstrip Antenna Design
A Combied Cotiuous/Biary Geetic Algorithm for Microstrip Atea Desig Rady L. Haupt The Pesylvaia State Uiversity Applied Research Laboratory P. O. Box 30 State College, PA 16804-0030 [email protected] Abstract:
Capacity of Wireless Networks with Heterogeneous Traffic
Capacity of Wireless Networks with Heterogeeous Traffic Migyue Ji, Zheg Wag, Hamid R. Sadjadpour, J.J. Garcia-Lua-Aceves Departmet of Electrical Egieerig ad Computer Egieerig Uiversity of Califoria, Sata
Malicious Node Detection in Wireless Sensor Networks using Weighted Trust Evaluation
Malicious Node Detectio i Wireless Sesor Networks usig Weighted Trust Evaluatio Idris M. Atakli, Hogbig Hu, Yu Che* SUNY Bighamto Bighamto, NY 1392, USA {iatakli1, hhu1, yche}@bighamto.edu Wei-Shi Ku Aubur
FACIAL EXPRESSION RECOGNITION BASED ON CLOUD MODEL
FACIAL EXPRESSION RECOGNITION BASED ON CLOUD MODEL Hehua Chi a, Liahua Chi b *, Meg Fag a, Juebo Wu c a Iteratioal School of Software, Wuha Uiversity, Wuha 430079, Chia - [email protected] b School of Computer
Optimize your Network. In the Courier, Express and Parcel market ADDING CREDIBILITY
Optimize your Network I the Courier, Express ad Parcel market ADDING CREDIBILITY Meetig today s challeges ad tomorrow s demads Aswers to your key etwork challeges ORTEC kows the highly competitive Courier,
