%PDF-1.4 % 1 0 obj << /S /GoTo /D (section*.2) >> endobj 4 0 obj (Contents) endobj 5 0 obj << /S /GoTo /D (chapter.1) >> endobj 8 0 obj (Getting Started with AD Model Builder) endobj 9 0 obj << /S /GoTo /D (section.1.1) >> endobj 12 0 obj (What are nonlinear statistical models?) endobj 13 0 obj << /S /GoTo /D (section.1.2) >> endobj 16 0 obj (Installing the software) endobj 17 0 obj << /S /GoTo /D (section.1.3) >> endobj 20 0 obj (The sections in an AD Model Builder \040tpl file) endobj 21 0 obj << /S /GoTo /D (section.1.4) >> endobj 24 0 obj (The original AD Model Builder examples) endobj 25 0 obj << /S /GoTo /D (section.1.5) >> endobj 28 0 obj (Example 1: linear least squares) endobj 29 0 obj << /S /GoTo /D (section.1.6) >> endobj 32 0 obj (The data section) endobj 33 0 obj << /S /GoTo /D (section.1.7) >> endobj 36 0 obj (The parameter section) endobj 37 0 obj << /S /GoTo /D (section.1.8) >> endobj 40 0 obj (The procedure section) endobj 41 0 obj << /S /GoTo /D (section.1.9) >> endobj 44 0 obj (The preliminary calculations section) endobj 45 0 obj << /S /GoTo /D (section.1.10) >> endobj 48 0 obj (The use of loops and element-wise operations) endobj 49 0 obj << /S /GoTo /D (section.1.11) >> endobj 52 0 obj (The default output from AD Model Builder) endobj 53 0 obj << /S /GoTo /D (section.1.12) >> endobj 56 0 obj (Robust nonlinear regression with AD Model Builder) endobj 57 0 obj << /S /GoTo /D (section.1.13) >> endobj 60 0 obj (Modifying the model to use robust nonlinear regression) endobj 61 0 obj << /S /GoTo /D (section.1.14) >> endobj 64 0 obj (Chemical engineering: \040a chemical kinetics problem) endobj 65 0 obj << /S /GoTo /D (section.1.15) >> endobj 68 0 obj (Financial Modelling: a generalized autoregressive conditional heteroskedasticity or garch model) endobj 69 0 obj << /S /GoTo /D (section.1.16) >> endobj 72 0 obj (Carrying out the minimization in a number of phases) endobj 73 0 obj << /S /GoTo /D (section.1.17) >> endobj 76 0 obj (Natural resource management: the Schaeffer-Pella-Tomlinson Model) endobj 77 0 obj << /S /GoTo /D (section.1.18) >> endobj 80 0 obj (Bayesian considerations in the Pella-Tomlinson Model) endobj 81 0 obj << /S /GoTo /D (section.1.19) >> endobj 84 0 obj (Using functions to improve code organization) endobj 85 0 obj << /S /GoTo /D (section.1.20) >> endobj 88 0 obj (A fisheries catch-at-age model) endobj 89 0 obj << /S /GoTo /D (section.1.21) >> endobj 92 0 obj (Bayesian inference and the profile likelihood) endobj 93 0 obj << /S /GoTo /D (section.1.22) >> endobj 96 0 obj (Saving the output from profile likelihood to use as starting values for mcmc analysis) endobj 97 0 obj << /S /GoTo /D (section.1.23) >> endobj 100 0 obj (The profile likelihood calculations) endobj 101 0 obj << /S /GoTo /D (section.1.24) >> endobj 104 0 obj (Modifying the profile likelihood approximation procedure) endobj 105 0 obj << /S /GoTo /D (section.1.25) >> endobj 108 0 obj (Changing the default file names for data and parameter input) endobj 109 0 obj << /S /GoTo /D (section.1.26) >> endobj 112 0 obj (Using the subvector operation to avoid writing loops) endobj 113 0 obj << /S /GoTo /D (section.1.27) >> endobj 116 0 obj (The use of higher-dimensional arrays) endobj 117 0 obj << /S /GoTo /D (section.1.28) >> endobj 120 0 obj (The TOP\137OF\137MAIN section) endobj 121 0 obj << /S /GoTo /D (section.1.29) >> endobj 124 0 obj (The GLOBALS\137SECTION) endobj 125 0 obj << /S /GoTo /D (section.1.30) >> endobj 128 0 obj (The BETWEEN\137PHASES\137SECTION) endobj 129 0 obj << /S /GoTo /D (chapter.2) >> endobj 132 0 obj (Markov Chain Simulation) endobj 133 0 obj << /S /GoTo /D (section.2.1) >> endobj 136 0 obj (Introduction to the Markov Chain Monte Carlo Approach in Bayesian Analysis) endobj 137 0 obj << /S /GoTo /D (section.2.2) >> endobj 140 0 obj (Reading AD Model Builder binary files) endobj 141 0 obj << /S /GoTo /D (section.2.3) >> endobj 144 0 obj (Convergence diagnostics for mcmc analysis) endobj 145 0 obj << /S /GoTo /D (chapter.3) >> endobj 148 0 obj (A Forestry Model: Estimating the Size Distribution of Wildfires) endobj 149 0 obj << /S /GoTo /D (section.3.1) >> endobj 152 0 obj (Model description) endobj 153 0 obj << /S /GoTo /D (section.3.2) >> endobj 156 0 obj (The numerical integration routine) endobj 157 0 obj << /S /GoTo /D (section.3.3) >> endobj 160 0 obj (Using the ad\137begin\137funnel routine to reduce the amount of temporary storage required) endobj 161 0 obj << /S /GoTo /D (section.3.4) >> endobj 164 0 obj (Effect of the accuracy switch on the running time for numerical integration) endobj 165 0 obj << /S /GoTo /D (section.3.5) >> endobj 168 0 obj (A comparison with Splus for the forestry model) endobj 169 0 obj << /S /GoTo /D (chapter.4) >> endobj 172 0 obj (Economic Models: Regime Switching) endobj 173 0 obj << /S /GoTo /D (section.4.1) >> endobj 176 0 obj (Analysis of economic data from hamilton1989) endobj 177 0 obj << /S /GoTo /D (section.4.2) >> endobj 180 0 obj (The code for Hamilton's fourth-order autoregressive model) endobj 181 0 obj << /S /GoTo /D (section.4.3) >> endobj 184 0 obj (Results of the analysis) endobj 185 0 obj << /S /GoTo /D (section.4.4) >> endobj 188 0 obj (Extending Hamilton's model to a fifth-order autoregressive process) endobj 189 0 obj << /S /GoTo /D (chapter.5) >> endobj 192 0 obj (Econometric Models: Simultaneous Equations) endobj 193 0 obj << /S /GoTo /D (section.5.1) >> endobj 196 0 obj (Simultaneous equations models) endobj 197 0 obj << /S /GoTo /D (section.5.2) >> endobj 200 0 obj (Full information maximum likelihood \(fiml\)) endobj 201 0 obj << /S /GoTo /D (section.5.3) >> endobj 204 0 obj (Concentrating out D for the fiml) endobj 205 0 obj << /S /GoTo /D (section.5.4) >> endobj 208 0 obj (Evaluating the model's performance) endobj 209 0 obj << /S /GoTo /D (section.5.5) >> endobj 212 0 obj (Results of \(fiml\) for unconstrained D) endobj 213 0 obj << /S /GoTo /D (section.5.6) >> endobj 216 0 obj (Results of \(fiml\) for constrained D) endobj 217 0 obj << /S /GoTo /D (section.5.7) >> endobj 220 0 obj (Code for \(fiml\) for constrained D) endobj 221 0 obj << /S /GoTo /D (chapter.6) >> endobj 224 0 obj (Truncated Regression) endobj 225 0 obj << /S /GoTo /D (section.6.1) >> endobj 228 0 obj (Truncated linear regression) endobj 229 0 obj << /S /GoTo /D (section.6.2) >> endobj 232 0 obj (The AD Model Builder truncated regression program) endobj 233 0 obj << /S /GoTo /D (chapter.7) >> endobj 236 0 obj (Multivariate garch) endobj 237 0 obj << /S /GoTo /D (section.7.1) >> endobj 240 0 obj (Formulation of the varma garch process) endobj 241 0 obj << /S /GoTo /D (section.7.2) >> endobj 244 0 obj (Setting a value for 1) endobj 245 0 obj << /S /GoTo /D (section.7.3) >> endobj 248 0 obj (Ensuring that the t are positive definite) endobj 249 0 obj << /S /GoTo /D (section.7.4) >> endobj 252 0 obj (Missing data) endobj 253 0 obj << /S /GoTo /D (section.7.5) >> endobj 256 0 obj (The likelihood function) endobj 257 0 obj << /S /GoTo /D (section.7.6) >> endobj 260 0 obj (Model selection) endobj 261 0 obj << /S /GoTo /D (section.7.7) >> endobj 264 0 obj (The Box-Ljung statistic) endobj 265 0 obj << /S /GoTo /D (section.7.8) >> endobj 268 0 obj (Analysis of simulated data) endobj 269 0 obj << /S /GoTo /D (section.7.9) >> endobj 272 0 obj (Analysis of real data) endobj 273 0 obj << /S /GoTo /D (subsection.7.9.1) >> endobj 276 0 obj (Model Parameters log-likelihood directory p,q,r,s.) endobj 277 0 obj << /S /GoTo /D (subsection.7.9.2) >> endobj 280 0 obj (Ljung-Box statistic \(chi2 with 10 degrees of freedom\).) endobj 281 0 obj << /S /GoTo /D (section.7.10) >> endobj 284 0 obj (Input format) endobj 285 0 obj << /S /GoTo /D (section.7.11) >> endobj 288 0 obj (Output files) endobj 289 0 obj << /S /GoTo /D (section.7.12) >> endobj 292 0 obj (The code for the bekkgarch model) endobj 293 0 obj << /S /GoTo /D (chapter.8) >> endobj 296 0 obj (The Kalman Filter) endobj 297 0 obj << /S /GoTo /D (section.8.1) >> endobj 300 0 obj (The Kalman filter) endobj 301 0 obj << /S /GoTo /D (section.8.2) >> endobj 304 0 obj (Equations for the Kalman filter) endobj 305 0 obj << /S /GoTo /D (section.8.3) >> endobj 308 0 obj (Parameterizing the covariance matrix parameterizations) endobj 309 0 obj << /S /GoTo /D (chapter.9) >> endobj 312 0 obj (Applying the Laplace Approximation to the Generalized Kalman Filter: with an Application to Stochastic Volatility Models) endobj 313 0 obj << /S /GoTo /D (section.9.1) >> endobj 316 0 obj (Parameter estimation) endobj 317 0 obj << /S /GoTo /D (section.9.2) >> endobj 320 0 obj (The stochastic volatility model) endobj 321 0 obj << /S /GoTo /D (section.9.3) >> endobj 324 0 obj (The data) endobj 325 0 obj << /S /GoTo /D (section.9.4) >> endobj 328 0 obj (The results) endobj 329 0 obj << /S /GoTo /D (chapter.10) >> endobj 332 0 obj (Using Vectors of Initial Parameter Types) endobj 333 0 obj << /S /GoTo /D (chapter.11) >> endobj 336 0 obj (Creating Dynamic Link Libraries with AD Model Builder) endobj 337 0 obj << /S /GoTo /D (section.11.1) >> endobj 340 0 obj (Compiling the code to produce a dll) endobj 341 0 obj << /S /GoTo /D (section.11.2) >> endobj 344 0 obj (The Splus objects) endobj 345 0 obj << /S /GoTo /D (section.11.3) >> endobj 348 0 obj (Debugging the dlls) endobj 349 0 obj << /S /GoTo /D (section.11.4) >> endobj 352 0 obj (Understanding what is being passed to the dll) endobj 353 0 obj << /S /GoTo /D (section.11.5) >> endobj 356 0 obj (Passing strings from Splus to a dll) endobj 357 0 obj << /S /GoTo /D (section.11.6) >> endobj 360 0 obj (A mixture of two bivariate normal distributions) endobj 361 0 obj << /S /GoTo /D (section.11.7) >> endobj 364 0 obj (Interpretation of the parameter estimates) endobj 365 0 obj << /S /GoTo /D (chapter.12) >> endobj 368 0 obj (Command Line Options) endobj 369 0 obj << /S /GoTo /D (chapter.13) >> endobj 372 0 obj (Writing Adjoint Code) endobj 373 0 obj << /S /GoTo /D (section.13.1) >> endobj 376 0 obj (The necessity for adjoint code) endobj 377 0 obj << /S /GoTo /D (section.13.2) >> endobj 380 0 obj (Writing adjoint code: a simple case) endobj 381 0 obj << /S /GoTo /D (section.13.3) >> endobj 384 0 obj (Debugging adjoint code: a simple case) endobj 385 0 obj << /S /GoTo /D (section.13.4) >> endobj 388 0 obj (Adjoint code for more than one independent variable) endobj 389 0 obj << /S /GoTo /D (section.13.5) >> endobj 392 0 obj (Structured calculation of derivatives in adjoint code) endobj 393 0 obj << /S /GoTo /D (section.13.6) >> endobj 396 0 obj (General adjoint code) endobj 397 0 obj << /S /GoTo /D (chapter.14) >> endobj 400 0 obj (Truncated Regression) endobj 401 0 obj << /S /GoTo /D (section.14.1) >> endobj 404 0 obj (Truncated linear regression) endobj 405 0 obj << /S /GoTo /D (section.14.2) >> endobj 408 0 obj (The AD Model Builder truncated regression program) endobj 409 0 obj << /S /GoTo /D (chapter.15) >> endobj 412 0 obj (All the Functions in AD Model Builder) endobj 413 0 obj << /S /GoTo /D (section.15.1) >> endobj 416 0 obj (Naming conventions for documenting functions) endobj 417 0 obj << /S /GoTo /D (section.15.2) >> endobj 420 0 obj (Mathematical functions) endobj 421 0 obj << /S /GoTo /D (section.15.3) >> endobj 424 0 obj (Operations on arrays) endobj 425 0 obj << /S /GoTo /D (subsection.15.3.1) >> endobj 428 0 obj (Element-wise operations) endobj 429 0 obj << /S /GoTo /D (section.15.4) >> endobj 432 0 obj (The identity matrix function identity\137matrix) endobj 433 0 obj << /S /GoTo /D (section.15.5) >> endobj 436 0 obj (Probability densities and related functions:poisson negative binomial cauchy) endobj 437 0 obj << /S /GoTo /D (section.15.6) >> endobj 440 0 obj (The operations det inv norm norm2 min max sum) endobj 441 0 obj << /S /GoTo /D (section.15.7) >> endobj 444 0 obj (Eigenvalues and eigenvectors of a symmetric matrix) endobj 445 0 obj << /S /GoTo /D (section.15.8) >> endobj 448 0 obj (The Choleski decomposition of a positive definite symmetric matrix) endobj 449 0 obj << /S /GoTo /D (section.15.9) >> endobj 452 0 obj (Solving a system of linear equations) endobj 453 0 obj << /S /GoTo /D (section.15.10) >> endobj 456 0 obj (Methods for filling arrays and matrices) endobj 457 0 obj << /S /GoTo /D (section.15.11) >> endobj 460 0 obj (Methods for filling arrays and matrices with random numbers) endobj 461 0 obj << /S /GoTo /D (section.15.12) >> endobj 464 0 obj (Methods for obtaining shape information from containers) endobj 465 0 obj << /S /GoTo /D (section.15.13) >> endobj 468 0 obj (Methods for extracting from arrays and matrices) endobj 469 0 obj << /S /GoTo /D (section.15.14) >> endobj 472 0 obj (Accessing subobjects of higher-dimensional arrays) endobj 473 0 obj << /S /GoTo /D (section.15.15) >> endobj 476 0 obj (Sorting vectors and matrices) endobj 477 0 obj << /S /GoTo /D (section.15.16) >> endobj 480 0 obj (Statistical functions) endobj 481 0 obj << /S /GoTo /D (section.15.17) >> endobj 484 0 obj (The random number generator class) endobj 485 0 obj << /S /GoTo /D (section.15.18) >> endobj 488 0 obj (The adstring class operations) endobj 489 0 obj << /S /GoTo /D (section.15.19) >> endobj 492 0 obj (Miscellaneous functions) endobj 493 0 obj << /S /GoTo /D (chapter.16) >> endobj 496 0 obj (Miscellaneous and Advanced Features of AD Model Builder) endobj 497 0 obj << /S /GoTo /D (section.16.1) >> endobj 500 0 obj (Using strings and labels in the tpl file) endobj 501 0 obj << /S /GoTo /D (section.16.2) >> endobj 504 0 obj (Using other class libraries in AD Model Builder programs) endobj 505 0 obj << /S /GoTo /D (section.16.3) >> endobj 508 0 obj (Using control files for bounded parameters) endobj 509 0 obj << /S /GoTo /D (chapter.1) >> endobj 511 0 obj (The Regression Function) endobj 512 0 obj << /S /GoTo /D (chapter.2) >> endobj 514 0 obj (AD Model Builder Types) endobj 515 0 obj << /S /GoTo /D (chapter.3) >> endobj 517 0 obj (The Profile Likelihood) endobj 518 0 obj << /S /GoTo /D (chapter.4) >> endobj 520 0 obj (Concentrated Likelihoods) endobj 521 0 obj << /S /GoTo /D (section*.38) >> endobj 524 0 obj (References) endobj 525 0 obj << /S /GoTo /D (section*.39) >> endobj 528 0 obj (Index) endobj 529 0 obj << /S /GoTo /D [530 0 R /Fit ] >> endobj 533 0 obj << /Length 433 /Filter /FlateDecode >> stream xuRMo0+hv<>6bm@Bl~ǘm"${fF$fʮn<B5P( o9xfklsU洟auCHH .8H
,[qb endstream endobj 530 0 obj << /Type /Page /Contents 533 0 R /Resources 532 0 R /MediaBox [0 0 612 792] /Parent 540 0 R >> endobj 531 0 obj << /Type /XObject /Subtype /Image /Width 285 /Height 300 /BitsPerComponent 8 /ColorSpace /DeviceRGB /Length 27448 /Filter/FlateDecode /DecodeParms<> >> stream xwt\qg}o I`XDzW$[%q/q/s؉ǎ#;,V$J^NE]zHD%)͙C.j>;9`9Giٻw.]SVZ,Be_llhr5u/7?, bTgre1!d\6mb?3;vI~bBw[VSje?zM}i ؿ0 Pfr5}*G`[o1Okh564 A :+zfreg\ af}{{0Y \6uܱ^Q+IFll4SɕiS'WG*o3%A+.HV{4<`IVco3Sɕi;4/̧X&Wz{}jb?|gɕiK/kk8u~,ML3]7vnDH$a&Wf`%+ ಬG͔ere ;vwtS^ڞ渠uo@u@>R 8o~&W{gOKG%.^8b&W{gOK_X/t̓#&W׳yў/qERtٚ\]EmmGhY \]E#sj.a^;i&Wש