Cornell College - JBITTER - 131

Jeff Bitter Computer Practice and Perspectives, CSC131 Question 2.26A.I believe that a database containing the names of convicted shoplifters for reference to store owners that subscribe is not a problem. First of all, the shoplifters have alread

Cornell College - JBITTER - 131

CSC1131 Computing Practice and Perspectives Exam 1 September 10, 2005 _Jeffrey R. Bitter_ name The only software you may use while completing this exam is Microsoft Word and the only document you may open is this one. You may save this document to e

Cornell College - KCOX - 131

H ank A Cl own aron was 1 s in t he N e 8 years ol d gro A meri c when he p an L e l ague. ayed f or t h e I nd i anaes Brav on Bost onl y e by t h he had r t yea Aaron a at t h ui red ht . boug ves acq sed gri p en as t h t he Bra a rever w aron .

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics1. Review 2. Types of a time series 3. Information contained in a time series Review What is a time series? Examples Some R codesTypes of a time series A time series is discrete A time series is c

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics1. Review 2. Stationarity (continuing) 3. More Statistical Definitions Review Objectives of time series analysis Description (plotting,. . . ) Explanation (modelling) Prediction (forecasting) Control S

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. More Statistical Definitions 3. Time Series with Trend and Seasonal Components Review Weak stationarity (Second order): stationarity in mean and covariance Autocorrelation k = k /0 at lag k; a

UWO - SS - 3861

1Statistical Sciences 3861B1. Review Today's Topics2. Time Series with Trend and Seasonal Components 3. Wold Theorem and Linear Time Series Models 4. Spectral Analysis 5. Ch3: Stationary Nonseasonal Models Review Memory: M = t=- |t | M < - sh

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Ch3: Stationary Nonseasonal Models 3. Autoregressive Processes Review Time Series with Trend and Seasonal Components Use k Xt to remove (polynomial) trend Use dXt to remove seasonality Spect

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Autoregressive Processes Review Chapter 3 Main objects: stationary linear noseasonal processes Three models: AR, MA, ARMA {Xt} is AR(1): Xt = 1Xt1 + at, 2 where 1 is the AR parameter and {at

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Autoregressive Processes Review (x) = 0 is called the characteristic equation of AR(p) Box and Jenkins: AR(p) is stationary iff (<=>) all roots of (x) = 0 must fall outside of the unit circ

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Autoregressive Processes Review ACF of AR(1): Two approaches 1 Xt = 1 B at = (1 + 1B + 2B 2 + )at 1 1 2 Xt = at + 1at1 + 1at2 + 2 2 EXtXt+k => (k) = ak j=0 2j = k a/(1 2), k = 1 1 1 1

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Moving Average Processes Review Yule-Walter Equations: p = P pp ^ ^ -1 ^ Y-W estimator of AR parameters: p = P p p Find a proper p in AR(p) process Extend Y-W equations Let (r) = (k1, .

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Moving Average Processes 3. Autoregressive Moving Average (ARMA) Processes Review Xt is MA(q ) if 2 Xt = at 1at1 q atq , at W N (0, a) 1, . . . , q are the MA paramters E[Xt] = E[at]

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Autoregressive Moving Average (ARMA) Processes Review MA(q ) is invertible if all roots of the characteristic equation (B) = 0 are outside of the unit circle. MA(q ) is invertible => it is

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics1. Review 2. Autoregressive Moving Average (ARMA) Processes 3. Constrained Models of ARMA Models 4. Box-Cox TransformationReview Xt is ARMA(p, q ) if it can be represented as Xt = 1Xt-1 + + pXt-p +

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Homogeneous Nonstationarity 3. ACF of ARIMA Models Review Many time series, in particular econometric time series, are not stationary Why are they not stationary? E[Xt] is not a constant (n

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Two formulations of the ARIMA Process 3. Integrate moving average process 4. Summary of Chapter 4 Review The ARIMA(p, d, q ) model is defined as (B)(1 - B) Xt = (B)at or (B) Let Zt = Xt =

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Model identication 2. Modelling philosophies 3. Identication methods 4. ExamplesModel identication The process of choosing a proper model The rst step of a model construction Also the most dicult an

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Examples 3. Chapter 6: Parameter estimation 4. Yule-Walker estimator 5. Some estimation theory Review Model identification: The process of choosing a proper model The most difficult and impo

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Yule-Walker estimator 3. Some estimation theory 4. Eciency of estimatorsReview Chapter 6 will do Find a criteria to choose proper p and q Estimate 1, . . . , p and 1, . . . , q Estimate the

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Maximum likelihood estimation 3. Model discrimination using AIC 4. Examples of ARMA parameter estimation Review Sample mean estimation of : CLT =>n/4 Xn N (, /n) , = k=n/4|k| 1 nk

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Model discrimination using AIC 3. Examples of ARMA parameter estimation 4. Purposes of Chapter 7 5. Overfitting Review For time series, L() = f (x1, . . . , xN ) = f1(x1)f2(x2|x1) fN (xN

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Residuals 3. Tests on ARMA parameters 4. Whiteness tests Review AIC=Akaike Information Criterion ARMA(p,q) models: AIC = 2 ln(max L() + 2k, k = p + q + 1 + . ARIMA models: AIC =N N d (2 ln

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Minimum MSE forecasts Review Constant variance tests Constant variance of innovations => homoscedasticity Changing (conditional) variance of innovations => heteroscedasticity One of main pu

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. Nonzero mean parameter in time series 3. Minimum MSE forecasts Review Use linear predictor Xt+l = b0 + b1Xt + + btX1 Use minimum MSE to nd b0, b1, . . . , bt? One-step prediction of an

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Review 2. Minimum MSE forecasts Review Time series with nonzero mean parameter : centering and estimation Model discrimination for an AR(p) model Fit a time series X1, . . . , XN with an AR(p) model

UWO - SS - 3861

1Statistical Sciences 3861BTodays Topics 1. Review 2. ARMA forecasts 3. ARIMA forecasts Review2 2 2 E(Xt+l Xt+l )2 = (1 + 1 + + l1)a E(Xt+1 Xt+1)2 = 2 a E(Xt+l Xt+l )2 Rules of prediction2 2 (1 + 1 + 2 + )a = V ar(Xt) 2 Xt+l

UWO - SS - 3861

1Statistical Sciences 361BToday's Topics 1. Review 2. ARIMA forecasts 3. Summary of Chapter 8 4. Purposes of Chapter 12 Review Apply the rules of prediction to l-step prediction of ARMA(p,q) in MA() form and in the original form Apply the rules

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Seasonal ARIMA models Seasonal ARIMA models ARIMA(p, d, q ) model: (B)(1 - B)dXt = (B)at (B) = 1 - 1B - - pB p (B) = 1 - 1B - - q B q (1 - B)d can remove a polynomial trend Seasonal AR ope

UWO - SS - 3861

1Statistical Sciences 3861BToday's Topics 1. Chapters 3, 4, 5 2. Chapter 6 3. Chapter 7 4. Chapter 8 5. Chapter 12 Chapters 3, 4, 5 AR(p), MA(q ), and ARMA(p, q) Stationarity and invertibility (Sample) ACF and PACF Yule-Walker equations Ho

Columbia - A - 6603

Fall 2006: PLAN A6603.001: Infrastructure Planning and International Economic Development Wednesday, 11 am-1 pm, Buell 300 Sumila Gulyani Email: sumila.gulyani@columbia.edu TA: Cuz Potter (jwp70@columbia.edu) Abstract Starting with old and new theori

Columbia - A - 6603

Columbia - A - 6603

Columbia - A - 6603

Pricing of infrastructure servicesNovember, 2006 Sumila GulyaniOutline1. 2. 3. 4.Definition and significance of user fees Tariff design: Theory & practice Case: Tariff reform & demand in Armenia Supply-side issues in improving cost recovery

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm Sumila GulyaniAssignment 1: Insights from the literatureHanded out: Sep. 6, 2006 Electronic submission due in Courseworks by 9 am on da

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm Sumila GulyaniAssignment 2: The Infrastructure "Business": Public Service Providers in Chicago/San Francisco/New YorkHanded out: Sept.

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm, Buell 300 Sumila GulyaniAssignment 1.3Handed out: October 23, 2006 Electronic submissions due on Monday, Oct 30, by 9 am Understanding

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm, Buell 300 Sumila GulyaniAssignment 1.4Handed out: November 2, 2006 Electronic submission due on Tuesday (by popular demand), Nov 14, b

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm, Buell 300 Sumila GulyaniAssignment 1.5Handed out: November 15, 2006 Electronic submissions due on Tuesday, November 21, by 9 am Infras

Columbia - A - 6603

Fall 2006: PLAN A6603.001Infrastructure Planning and International Economic DevelopmentWednesday, 11 am-1 pm, Buell 300 Sumila GulyaniAssignment 3Handed out: November 22, 2005 Electronic submissions due on Tuesday, Dec. 5 by 9 am In-class prese

Columbia - V - 3565

Columbia - A - 3565

University of Montana - MBA - 600

The University of Montana GRADUATE DEGREES ID or Social Security NumberAPPLICATION FOR GRADUATIONThis document must be approved and signed by your adviser before submitting the original and two copies to the Graduate School at leas

University of Hawaii - Hilo - EE - 693

The Tides of EDAAlberto Sangiovanni-VincentelliUniversity of California at BerkeleyAlberto bases this article on remarks from his invited keynote speech at the 40th Design Automation Conference. In that speech, he proposed a bold initiative in ele

University of Hawaii - Hilo - EE - 693

COVER FEATUREA Decade of Hardware/ Software CodesignHardware/software codesign has been a recognized research eld for about a decade. Within that time, it has moved from an emerging discipline to a mainstream technology.Wayne WolfPrinceton Univ

University of Hawaii - Hilo - EE - 693

COVER FEATURELeakage Current: Moores Law Meets Static PowerMicroprocessor design has traditionally focused on dynamic power consumption as a limiting factor in system integration. As feature sizes shrink below 0.1 micron, static power is posing ne

University of Hawaii - Hilo - EE - 693

Other aspectsCode compressionExtreme version of instruction encoding: Use variable-bit instructions. Generate encodings using compression algorithms. Generally takes longer to decode. Can result in performance, energy, code size improve

University of Hawaii - Hilo - EE - 693

44.1RISPP: Rotating Instruction Set Processing PlatformLars Bauer, Muhammad Shafique, Simon Kramer and Jrg HenkelUniversity of Karlsruhe, Chair for Embedded Systems, Karlsruhe, Germany {lars.bauer, shafique, henkel} @ informatik.uni-karlsruhe.de

University of Hawaii - Hilo - EE - 693

A Lock-Free Multiprocessor OS KernelHenry Massalin and Calton Pu Department of Computer Science Columbia University New York, NY 10027 Technical Report No. CUCS-005-91calton@cs.columbia.eduRevised June 19, 1991AbstractTypical shared-memory mul

University of Hawaii - Hilo - EE - 693

Adaptive Operating System Abstractions: A Case Study of Multiprocessor LocksBodhisattwa Mukherjee (bodhi@cc.gatech.edu) Karsten Schwan (schwan@cc.gatech.edu)GIT{CC{94/3910 June 1994AbstractOperating system kernels typically o er a xed and limi

University of Florida - CGS - 3220

CGS 3220 Lecture 1Introduction to Computer Aided ModelingInstructor: Brent RossenJason HillhouseSyllabus Prerequisites and ContactCourse Webpage:http:/www.cise.ufl.edu/~brossen/cgs3220 Announcements + Project Descriptions Copy of syllabus i

University of Florida - CGS - 3220

CGS 3220 Lecture 1Introduction to Computer Aided Modeling Instructor: Brent RossenJason Hillhouse Syllabus Prerequisites and ContactCourse Webpage: http:/www.cise.ufl.edu/~brossen/cgs3220 Announcements + Project Descriptions Copy of

University of Florida - CGS - 3220

CGS 3220 Lecture 10 Dynamic Rigid BodiesIntroduction to Computer Aided ModelingInstructor: Brent RossenOverviewCreating a Passive Rigid Body Creating an Active Rigid Body Adding a gravity field Simulating dynamics Setting rigid body attributes

University of Florida - CGS - 3220

CGS 3220 Lecture 11 Camera Animation, Rendering, and CompressionIntroduction to Computer Aided ModelingInstructor: Brent RossenOverviewThe Imperfect CameraMaya's Perfect Camera Imperfecting: Depth of Field and Lens Flares Film Camera Aperture:

University of Florida - CGS - 3220

CGS 3220 Lecture 12 NURBS Modeling (Beginning Character Modeling)Introduction to Computer Aided ModelingInstructor: Brent RossenOverviewManipulating NURBS Projecting a curve onto a surface Trim a surface Snap points to curves and isoparms Duplic

University of Florida - CGS - 3220

CGS 3220 Lecture 13 Polygonal Character ModelingIntroduction to Computer Aided ModelingInstructor: Brent RossenOverviewBox modeling Polygon proxy Mirroring Polygonal components Topology editing Procedural modeling attributes Changing edge norma

University of Florida - CGS - 3220

CGS 3220 Lecture 14 Polygonal TexturingIntroduction to Computer Aided ModelingInstructor: Brent RossenOverviewProjecting textures onto polygons Manipulating projections Using the UV Texture Editor Growing and reducing the current selection Assi

University of Florida - CGS - 3220

CGS 3220 Lecture 15 MEL Shelf Buttons Posing the SkeletonIntroduction to Computer Aided ModelingInstructor: Brent RossenOverview Script Editor and Posing Exploring the Script Editor MEL Shelf Buttons Posing the Skeleton Forward Kinematics

University of Florida - CGS - 3220

CGS 3220 Lecture 16 Character Skinning Painting WeightsIntroduction to Computer Aided ModelingInstructor: Brent RossenOverview Smooth Bind Skin WeightsCharacter SkinningSkinning: the process of connecting a character's meat to his bones.

University of Florida - CGS - 3220

CGS 3220 Lecture 17 Subdivision SurfacesIntroduction to Computer Aided ModelingInstructor: Brent RossenOverviewConverting from polygons to subdivision surfaces (sub-d) Modeling with sub-d using polygon proxy Adding detail Using creases Basic bu

University of Florida - CGS - 3220

Graham ClarkCGS 3220 Lecture 2Introduction to Computer Aided ModelingInstructor: Brent RossenLesson 1 Create a GarageOverviewSetting a new Maya Project Creating primitive objects Moving objects in 3d space Duplicating objects Changing the s