https://github.com/jimbrig/lossrunAnalyzer

R Package and Shiny App to Analyze Insurance Lossruns
https://github.com/jimbrig/lossrunAnalyzer

actuarial data-analysis data-mining data-science insurance r record-linkage risk-management shiny

Last synced: 26 days ago
JSON representation

R Package and Shiny App to Analyze Insurance Lossruns

• Host: GitHub
• URL: https://github.com/jimbrig/lossrunAnalyzer
• Owner: jimbrig
• Archived: true
• Created: 2020-01-10T20:01:18.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-01-10T20:52:20.000Z (over 4 years ago)
• Last Synced: 2024-05-02T03:08:54.745Z (2 months ago)
• Topics: actuarial, data-analysis, data-mining, data-science, insurance, r, record-linkage, risk-management, shiny
• Language: R
• Size: 11.7 KB
• Stars: 3
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.Rmd

Lists

• jimsghstars - jimbrig/lossrunAnalyzer - R Package and Shiny App to Analyze Insurance Lossruns (R)

README

        
---

output: github_document

---

```{r, include = FALSE}

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%"

)

```

# lossrunAnalyzer - Analyzing Insurance Claims Data

The goal of **lossrunAnalyzer** is to assist actuaries to quickly analyze, 

diagnose, and summarize lossruns containing individual claims data for 

property casualty insurance.

## Badges

[![Lifecycle: experimental](https://img.shields.io/badge/lifecycle-experimental-orange.svg)](https://www.tidyverse.org/lifecycle/#experimental)

[![Project Status: WIP](https://www.repostatus.org/badges/latest/wip.svg)](http://www.repostatus.org/#wip)

## Installation

You can install lossrunAnalyzer from [GitHub](https://github.com/) with:

``` r

# install.packages("devtools")

devtools::install_github("jimbrig2011/lossrunAnalyzer")

```

## Roadmap

The end-goal of **lossrunAnalyzer** is to provide support for the following: 

- Initial Reasonability Checks:

  + Unique claim ID

  + Paid + Case = Incurred

  + Totals = Sum of Splits

  + Report Date >= Loss Date

  + Field Consistency (i.e. States, Status, etc.)

  

- Possible Duplicate Detection

- Occurrence Grouping

- Adding "Working" Fields

  + Retentions / Limits / Deductibles

  + Various "Limited" Amounts

  + ALAE Treatments

  + Years (Policy, Accident, Report, Fiscal, Calendar)

  + Lags (Report, Close, Tenure)

  + Max IBNR's at various scenarios

  + Legal, Lost-Time, Indemnity Support

- Utilizing Lookup / Support Tables

- Record Linkage to Reduce Fuzzyness

- Merging Lossruns Across Multiple Evaluations

- Comparing Lossruns to Prior's and Addressing Various KPI's / Diagnostic Checks

- Anomaly / Outlier Detection

- Automating Development Comments as to why things changed

- Summarizing Data into Triangles

- Performing an AvE (Actual vs. Expected) Analysis Summary

- Checking for Dropped / Missing Claims or New Claims with Old Dates

- Tie Out to Exposures