ReSurv

ReSurv is an R software for predicting IBNR claims. The software includes tools for synthetic data generation, data pre-processing, hyperparameters tuning, model estimation and prediction.

The package is based on the approach illustrated in Hiabu M., Hofman E., and Pittarello G. (2023) and estimates feature dependent development factors using individual reserving data.

Available Machine Learning (ML) models

There is a one-to-one relationship between development factors and hazard rates (Hiabu et al. (2023)). The package implements extends the following machine learning algorithms for proportional hazard models:

• Cox model with splines (COX, Gray (1992)).

• Neural Networks (NN, Katzman et al. (2018)).

• eXtreme Gradient Boosting (XGB, Chen et al. (2016)).

ReSurv extends COX, NN, and XGB to account for ties in left-truncated and right-censored observations.

Installation

Developer Version

The developers version of the package can be installed from GitHub.

devtools::install_github('https://github.com/edhofman/ReSurv')

Python Dependencies

For using the NN models we suggest to install a virtual environment using

install_pyresurv()

The default name of the virtual environment is "pyresurv".

We then suggest to refresh the R session and to import the ReSurv package in R using

library(ReSurv)
reticulate::use_virtualenv("pyresurv")

Managing Multiple Package Dependencies

This section is taken from the guidelines of the R package reticulate for handling the case of multiple packages in your session that used isolated-package-environments. The most straightforward solution would be installing a dedicated environment for both.

envname <- "./venv"
ReSurv::install_pyresurv(envname = envname)
pysparklyr::install_pyspark(envname = envname)

References

• Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785-794).

• Gray, R. J. (1992). Flexible methods for analyzing survival data using splines, with applications to breast cancer prognosis. Journal of the American Statistical Association, 87(420), 942-951.

• Hiabu, M., Hofman, E., & Pittarello, G. (2023). A machine learning approach based on survival analysis for IBNR frequencies in non-life reserving. arXiv preprint arXiv:2312.14549.

• Snoek, J., Larochelle, H., & Adams, R. P. (2012). Practical bayesian optimization of machine learning algorithms. Advances in neural information processing systems, 25.

• Katzman, J. L., Shaham, U., Cloninger, A., Bates, J., Jiang, T., & Kluger, Y. (2018). DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network. BMC medical research methodology, 18, 1-12.