The hirem package

This is the source code for the hirem package, which is currently under development.

Installation

To install hirem from GitHub you will need devtools:

install.packages('devtools')
devtools::install_github('jonascrevecoeur/hirem')

Overview

The hirem package offers tools for implementing hierarchical reserving models, introduced in the paper Crevecoeur, J., Antonio, K., A generalized reserving model: bridging the gap between pricing and individual reserving, (2019).

The hirem package constructs hierarchical reserving models by sequentially adding new layers. An example of a three component hierarchical model:

require(hirem)
data("reserving_data")

model <- hirem(reserving_data %>% dplyr::filter(calendar_year <= 6)) %>%
  layer_glm('close', binomial(link = logit)) %>%
  layer_glm('payment', binomial(link = logit)) %>%
  layer_glm('size', Gamma(link = log),
            filter = function(data){data$payment == 1})
            
model <- fit(model,
             close = 'close ~ factor(development_year)',
             payment = 'payment ~ close + factor(development_year)',
             size = 'size ~ close + factor(development_year)')
            

The package is still under development. Currently supported layers are:

• layer_glm: Estimates a layer using a generalized linear model
• layer_gbm: Estimates a layer using a gradient boosting model

Supported distributions are:

• binomial
• gaussian
• gamma

To use the gamma distribution in layer_gbm you require an experimental version of the package gbm that implements the gamma distribution. See https://github.com/harrysouthworth/gbm.