close Warning: Can't synchronize with repository "(default)" (/var/svn/tolp does not appear to be a Subversion repository.). Look in the Trac log for more information.

Opened 14 years ago

Closed 14 years ago

#1037 closed task (fixed)

Qualitative Response Model

Reported by: Jorge Owned by: Víctor de Buen Remiro
Priority: highest Milestone: TOL Packages
Component: Math Version: 2.0.1
Severity: blocker Keywords: Probit, Logit
Cc:

Description

Implement a package to deal with regression models when the dependent variable is qualitative. This package will be called QltvRespModel.

Change History (20)

comment:1 Changed 14 years ago by Víctor de Buen Remiro

(In [2861]) Refs #1037
El uso del gradiente acelera notablemente el proceso

comment:2 Changed 14 years ago by Víctor de Buen Remiro

(In [2862]) Refs #1037

comment:3 Changed 14 years ago by Víctor de Buen Remiro

(In [2864]) Refs #1037

comment:4 Changed 14 years ago by Víctor de Buen Remiro

(In [2865]) Refs #1037
Nueva clase WgtBinReg (Weighted binary regression) de la que heredar WgtLogit y WgtProbit

comment:5 Changed 14 years ago by Víctor de Buen Remiro

(In [2867]) Refs #1037

comment:6 Changed 14 years ago by Jorge

(In [2868]) refs #1037, developing the gradient & hessian for the weighted logit

comment:7 Changed 14 years ago by Jorge

(In [2869]) refs #1037, W is diagonal

comment:8 Changed 14 years ago by Víctor de Buen Remiro

(In [2872]) Refs #1037
Renombrado

comment:9 Changed 14 years ago by Víctor de Buen Remiro

(In [2873]) Refs #1037
lapsus

comment:10 Changed 14 years ago by Víctor de Buen Remiro

(In [2911]) Refs #1037

comment:11 Changed 14 years ago by Víctor de Buen Remiro

(In [2917]) Refs #1037
Comments

comment:12 Changed 14 years ago by Víctor de Buen Remiro

(In [2919]) Refs #1037
Adding Class @WgtProbit

comment:13 Changed 14 years ago by Víctor de Buen Remiro

(In [2923]) Refs #1037
Adding statistical information about max-likelihood and bayesian estiamtions

comment:14 Changed 14 years ago by Jorge

(In [2925]) refs #1037, more documentation about max-llk logit

comment:15 Changed 14 years ago by Víctor de Buen Remiro

(In [2929]) Refs #1045
Refs #1037

comment:16 Changed 14 years ago by Víctor de Buen Remiro

(In [2949]) Refs #1037
Se usa la estimación asintótica máximo-verosímil de la varianza para ajustar el tamaño de paso del simulador bayesiano

comment:17 Changed 14 years ago by Víctor de Buen Remiro

(In [2950]) Refs #1037
Se usa la estimación asintótica máximo-verosímil de la varianza para ajustar el tamaño de paso del simulador bayesiano

comment:18 Changed 14 years ago by Víctor de Buen Remiro

(In [2951]) Refs #1037
Se usa la estimación asintótica máximo-verosímil de la varianza (diagonal, sin covarianzas) para ajustar el tamaño de paso del simulador bayesiano

comment:19 Changed 14 years ago by Víctor de Buen Remiro

(In [2952]) Refs #1037
Se usa la estimación asintótica máximo-verosímil de la matriz de covarianzas (con simetría forzada para eviatra problemas numéricos) para ajustar el tamaño de paso del simulador bayesiano

comment:20 Changed 14 years ago by Víctor de Buen Remiro

Resolution: → fixed
Status: new → closed
Note: See TracTickets for help on using tickets.