- .C
- .Call
- .External
- .Fortran
- 1D minimizer, see minimization
- one dimensional
- 1D roots, see roots
- one dimensional
- 6–12 potential, see Lennard-Jones potential
-
- acknowledgements
- active set
- see constraint
- active set
- AD, see automatic differentiation
- AD Model Builder, see ADMB
- adagio
- ADMB
- phases (fixed parameters)
- alabama
- algorithm
- controls for
- Anstee R
- approximate and descend
- approximation problem
- argument
- auglag()
- automatic differentiation,
Automatic Differentiation Model Builder, see ADMB CRAN
- axial search
-
- barrier function
- logarithmic
- BASIC
- Bates D
- Bayesian inference using Gibbs sampling, see BUGS Markow chain Monte Carlo,
- BB
- bbmle
- Beale–Sorenson method
- bell-shaped curve
- BFGS
- BFGS update
- binomial theorem
- bisection
- bobyqa()
- Bolker B
- Bolstad J
- Borchers H W
- bounds, see constraint
- box constraint, see constraint
- bounds
- Box G.
- Brent R
- Broyden–Fletcher–Goldfarb–Shanno update, see BFGS,
- BUGS
- byte-code compiler
-
- C
- Canada Interest Act
- Cauchy A-L
- centering
- CG
- CG, see conjugate gradients
- checking computations
- checks
- software
- classification
- problems
- Cobb–Douglas production function
- coef()
- codecoef()
- coefficient, see parameters
- coefficients
- extracting from solution
- COIN-OR
- compiler
- Comprehensive R Archive Network
- computable function
- computational efficiency, see efficiency
- computational
- computational performance, see performance
- confidentiality problem
- conjugate gradients
- Conn A
- constraint
- active set
- bounds
- equality
- elimination
- penalty method
- for L1 regression
- general
- inequality
- linear
- nonlinear
- nonnegativity
- parameter elimination
- satisfied
- unstated
- violated
- violation
- constraint-based optimization, see mathematical programming,
- constrOptim.nl()
- contour plot
- control defaults
- controls
- optimizer
- convergence
- quadratic
- convergence test
- relative offset criterion
- convergence tolerance
- convex optimization
- CPLEX
- CRAN
- cross-entropy method
- crossprod()
- curve()
-
- Dantzig G
- data
- exogenous
- global
- Davidon–Fletcher–Powell update, see DFP
- default control settings
- Dekker T
- Dennis J
- DEoptim
- derivative
- analytic
- approximation
- with constraint
- automatic
- backward
- central
- complex step
- example code
- forward
- incorrect
- methods
- comparison
- numerical
- Richardson extrapolation
- strategy
- symbolic
- tools
- derivative-free method
- deSolve
- deviance()
- dfoptim
- d foptim
- DFP
- diet problem
- differential equation models
- differential evolution
- differentiation
- symbolic
- tools
- digit cancellation
- dimension
- reduction by constraint
- dput()
- Duggleby R
- dyn.load()
- dyn.load
-
- effective monthly interest rate
- efficiency
- computational
- eigen()
- Eisenberg M
- equality
- test for
- equality constraint
- equation
- nonlinear
- one variable
- system of
- error
- sign of result
- evolutionary operation
- execution profile, see profiling
- execution time, see timing
- exogenous data
- expit
- expm
- exponential function
- external tools
- linking to
-
- failure of computations
- feasible region
- fitdistr()
- fitdistrplus
- fixed parameters, see masks, see masks constraint276
- Fletcher R
- Fletcher–Reeves method
- Fortran
- free libre software
- free parameters
- function
- computable
- external language compiled
- undefined
- example
- function names
- confusion
- confusion of
- duplicate
-
- GA
- gaoptim
- Gauss C-F
- Gauss–Newton method
- Gaussian density
- gaussNewton()
- Gay D
- generalized inverse
- generalized Rosenbrock function, see Rosenbrock function,
- genetic optimization methods
- GenSA
- Geyer C
- global minimum
- global optimization
- gnm
- Gnu Linear Programming Kit, see Rglpk
- goal of book
- Golden-section search
- Golub G
- gradient
- gradient method
- GrassmannOptim
- grid search
- grofit
- Grothendieck G
-
- Hansen E
- Hartley's method
- Hassan problem
- Hessian
- eigenvalues
- positive definite
- reporting
- singular
- when bounds active
- Hessian matrix, see Hessian
- heuristic search
- hjk()
- hjkb()
- Hobbs D
- Hobbs weed infestation problem
- Hompack
- Hooke and Jeeves method
- Huet S
- hyperbolic transformation
-
- inadmissible inputs
- inadmissible results
- indicator vector
- inexact Newton method
- infeasible region
- infeasible starting point
- initial point, see starting point
- integer programming
- interface
- software, see software interfaces
- to optimization
- interior point optimizer
- interpolation
- inverse
- interval analysis
- inverse interpolation, see interpolation
- inverse Brent R
- ipop()
- ipopt()
- ipoptr
- is.loaded()
- iterations
- iterative algorithm
-
- Jacobian
- approximation
- singular
- singular values
- Jacobian matrix, see Jacobian
- JAGS
- Joe H
- Johnson S
- Johnson S G,
- Karush W
- Kaufman L
- Kelley T
- kernlab
- KKT conditions
- see optimality conditions
- knitr
- Koenker R
- Kovalchik S
- Krogh F
- Kuhn H W
-
- L-BFGS-B
- L-infinity regression, see minimax regression
- L1 norm
- L1 regression
- L1linreg()
- L2 norm
- LAD regression, see L1 regression
- Lanczos C
- Least Absolute Deviations regression, see L1 regression
- least squares
- for nonlinear equations
- linear constraints
- nonlinear, see nonlinear least squares
- legal issues for R software
- Legendre A-M
- Lemoine N
- Lennard–Jones potential
- Levenberg Marquardt stabilization
- LEXPIT model
- likelihood
- likelihood function
- gradient
- limited memory method
- Limited-memory BFGS, see L-BFGS-B
- line-search
- Linear programming
- for L1 regression
- linearity
- partial
- linking
- to external tools
- via files
- linprog
- Linux
- local minimum
- log
- of objective function
- log likelihood
- log scaling
- log transformation
- logarithmic barrier function
- logistic growth model, see model, logistic growth
- logistic regression
- logit
- loops
- for
- while
- avoiding
- LP, see Linear Programming
- lpSolve
- LSODA differential equation solver
- Lyness J
-
- Maechler M
- Markow chain Monte Carlo
- Marquardt D
- Marquardt method
- masks (fixed parameters)
- MASS
- mathematical programming
- Matrix
- matrix
- $n=1$
- matrix multiplication
- implicit
- maximization, see minimization224
- maximum absolute error
- maximum absolute residual
- see minimax regression
- maximum likelihood, see likelihood function143
- maxLik
- mcga
- measures of performance
- method
- change of
- controls for
- Michaelis-Menten model
- microbenchmark
- Microsoft Windows
- minima
- multiple, see multiple minima
- Minimax approximation
- minimax regression
- by linear programming
- minimization
- constrained
- for nonlinear equations
- gradient method
- one dimensional
- one parameter
- unconstrained
- minimum
- from root-finding
- one dimensional
- minimum sum of absolute residuals, see L1 regression
- minpack.lm
- minqa
- missing data
- leading to termination
- mixed effect model
- mixed integer programming
- mle2()
- model
- cholera outbreak
- LEXPIT
- linear
- logistic growth
- mixed effect
- SIWR
- model2grfun()
- model2jacfun()
- model2resfun()
- model2ssfun()
- modgr()
- modss()
- monitor progress
- mortgage rate
- Canada
- MOSEK
- Mullen K
- multinomial maximum likelihood problem
- multiple exponentials problem
- multiple minima
- multiple starts
- Murdoch D
-
- names for quantities
- confusion
- Nash J C
- negative log likelihood, see likelihood function
- Nelder J
- Nelder–Mead method
- NEOS server
- newt1d()
- Newton I
- Newton's method
- Newton's method
- difficulties
- one dimensional
- newuoa()
- NISTnls
- nleqslv
- nlfb()
- nlm()
- nlme
- nlminb()
- nlmrt
- nlmrt the model expression and generate R functions:,
- nlmrt
- NLopt
- nloptr
- nloptwrap
- NLP, see nonlinear programming
- nlrq()
- nls()
- plinear algorithm
- nls()
- nls() ‘port’ algorithm
- concerns with
- example
- structure of solution
- nls.lm()
- nls2
- nlsLM()
- nlstools
- nlsystemfit()
- nlxb()
- nlxb
- nmk()
- nmkb()
- nnet()
- nnls
- nnls, see nonnegative least squares
- Nocedal J
- nonlinear
- nonlinear equations
- by minimization
- by nonlinear least squares
- nonlinear function minimization
- nonlinear least squares
- ancillary tools
- bounds
- extreme example
- for nonlinear equations
- Jacobian
- modeling expression
- conversion to function
- relative offset termination test
- residual function
- self starting models
- small sumsquares termination test
- starting parameters
- strategies
- zero residual problem
- nonlinear model
- generalized
- nonlinear programming
- packages
- nonlinear regression
- nonlinearity
- nonnegative least squares
- normal density, see Gaussian density
- nsga2R
- numDeriv
-
- obfuscating published statistics
- objective function
- Fortran
- fuzzy
- imprecise
- imprecisely evaluated
- log scale
- noisy
- properties
- quadratic
- structure
- Oleary D
- one-parameter roots, see roots
- one dimensional
- open source software
- OpenBUGS
- opposite sign requirement
- optextras
- optim()
- optim() BFGS
- optimality conditions
- optimization
- convex
- general constrained
- global
- stochastic
- trajectory
- optimize()
- example
- optimizer project (R-forge)
- optimizer controls
- optimx
- optplus
- outer product
-
- package
- upgrade
- package stats
- pander
- parallel tempering
- parameter
- extracting from solution
- free
- parameter elimination
- parma
- partial linearity
- particle swarm algorithms
- paster regrowth problem
- pattern search
- PBSadmb
- penalty
- sequential increasing
- penalty methods
- ill-conditioning
- Pereyra V
- performance
- improvement
- measures
- performance calibration
- example
- performance tuning, see tuning
- pkgucminf
- Polak–Ribiére method
- polynom
- polynomial approximation
- polynomial roots
- polyroot()
- polytope
- Powell M J D
- pracma
- predicted variable
- predictor variable
- premature termination
- Price K
- print()
- probability collectives
- /proc/cpuinfo
- /proc/meminfo
- production function
- profiling
- profr
- programming style
- progress monitoring
- project optimizer (R-forge)
- projection
- proprietary software
-
- QP, see quadratic programming
- quadprog
- quadratic convergence, see convergence
- quadratic
- Quadratic Programming
- quantile regression
- nonlinear
- quantreg
- quantreg Koenker R
- quantum annealing
- quasi-Newton method
-
- R
- R
- base system
- R formula object
- R-forge
- R2admb
- random search
- Ratkowsky D
- Rayleigh quotient
- timings
- rbenchmark
- Rcgmin
- R cgmin
- RcplexCPLEX
- RcppDE
- Rdonlp2
- reactive search optimization
- regression
- versus least squares
- relative offset convergence test
- see convergence test
- relative offset criterion
- reliability
- Remez algorithm
- reparameterization
- reportr
- require()
- residuals
- results
- inadmissible
- output of
- results output
- return code
- rgenoud
- Rglpk
- rgp rgp
- Rmalschains
- Rmosek
- Rmpfr
- robustness
- of nonlinear least squares methods
- of optimization
- root1d()
- rootoned (R -forge)
- roots
- bracketing
- complex
- example
- failure
- methods
- multiple
- one dimensional
- polynomial
- software requirements
- starting interval
- Rosenbrock function
- generalized
- Rprof()
- Rsolnp
- Rsymphony
- Rtools
- Runge–Kutta method
- Rust B
- Rvmmin
- R vmmin
-
- saddle point
- Sande G
- SANN
- SANN()
- scale()
- scaling
- log, see log scaling likelihood function143
- Schnabel R
- Seber G
- secant method
- second derivative matrix, see Hessian
- selfStart() methods
- sem
- separable least squares
- separable sums of squares
- setRNG
- settings
- default
- shadow price
- sign error
- simplex
- Simplex method (Dantzig)
- Simulated Annealing
- “singular gradient”
- singular Hessian, see Hessian
- singular
- singular values
- Jacobian
- singularity
- SIWR model
- small residuals
- smco
- software interfaces
- software structure
- solnp()
- solution
- acceptable
- failure
- inadmissible
- multiple
- test of
- unstated constraint
- unwanted
- user requirements on
- Solve.QP()
- solver
- commercial
- soma
- speedup, see tuning
- spg()
- square norm
- square transformation
- starting interval
- starting parameters, see starting point
- starting point
- infeasible
- multiple
- starting vector
- stats4
- steepest descents
- example
- stepsize
- too big
- stiff differential equation
- stochastic hill climbing
- stochastic optimization
- stochastic tunneling
- structural equations
- success indicator
- success–failure search
- sum of absolute errors
- summary()
- sumscale problem
- SUMT (sequential unconstrained minimization technique)
- symbolic differentiation
- symbolic differentiation, see differentiation
- symbolic
- symbolic maths
- system call
- system()
- system.time()
- systemfit
-
- taboo Search
- tanh transformation
- termination
- abnormal
- missing data
- termination test
- abnormal ending
- normal finish
- tests
- initial
- of solution HAI see solution
- test of
- of solutions
- performance loss due to
- software
- thin client
- ∼(tilde)
- timing
- calibration
- example
- improvement
- of computations
- tolerances
- tools
- Torczon V.
- transfinite()
- transformation
- affine
- hyperbolic
- log
- matrix
- square
- transformation tanh
- truncated Newton method
- trust
- trustOptim
- Tucker A W
- tuning, see performance
- two straight line problem
-
- ucminf
- UNCMIN
- undefined function, see function
- undefined
- uniroot
- unirootR()
- uniroot()
- upgrade package, see package
- upgrade
- Varadhan R
- variable
- dependent
- endogenous
- exogenous
- global
- independent
- predicted
- predictor
- variable metric method
- variable projection method
- VARPRO
- vectorization
- vectorized code
-
- Walster W
- Watson L
- Watts D
- weight loss example
- weighted least squares
- weighted nonlinear least squares
- weighted nonlinear regression
- weights
- Wickham H
- Wild C
- WinBUGS
- Windows (Microsoft), see Microsoft Windows
- WOBLRAMP function
- wrapnls()
-
- Ypma J
-
- zero
- test for
- zero divide
- zeroin()
..................Content has been hidden....................
You can't read the all page of ebook, please click
here login for view all page.