Analytic Insurance Rating Techniques
Automatically create a ggplot for objects obtained from univariate()
Set reference group to the group with largest exposure
Find active rows per date
Smooth coefficients in the model
Automatically create a summary for objects obtained from reduce()
Univariate analysis for discrete risk factors
Create new offset-term and new formula
Refitting Generalized Linear Models
Add predictions to a data frame
Automatically create a ggplot for objects obtained from bootstrap_rmse...
Automatically create a ggplot for objects obtained from check_residual...
Automatically create a ggplot for objects obtained from construct_tari...
Generalized additive model
Fit a distribution to truncated severity (loss) data
Create a histogram with outlier bins
Get model data
Performance of fitted GLMs
Split period to months
Root Mean Squared Error
Automatically create a ggplot for objects obtained from fit_gam()
Automatically create a ggplot for objects obtained from restrict_coef(...
Automatically create a ggplot for objects obtained from rating_factors...
Automatically create a ggplot for objects obtained from smooth_coef()
Automatically create a ggplot for objects obtained from fit_truncated_...
Bootstrapped RMSE
Check overdispersion of Poisson GLM
Check model residuals
Construct model points from Generalized Linear Model
Construct insurance tariff classes
Fisher's natural breaks classification
Include reference group in regression output
Reduce portfolio by merging redundant date ranges
Objects exported from other packages
Refitting Generalized Linear Models
Restrict coefficients in the model
Generate data from truncated gamma distribution
Generate data from truncated lognormal distribution
Functions to build, evaluate, and visualize insurance rating models. It simplifies the process of modeling premiums, and allows to analyze insurance risk factors effectively. The package employs a data-driven strategy for constructing insurance tariff classes, drawing on the work of Antonio and Valdez (2012) <doi:10.1007/s10182-011-0152-7>.
Useful links