Risk Evaluation in the Insurance Company Using REFII Model

Risk Evaluation in the Insurance Company Using REFII Model

Goran Klepac
DOI: 10.4018/978-1-4666-8473-7.ch038
(Individual Chapters)
No Current Special Offers


A business case describes a problem present in all insurance companies: portfolio risk evaluation. Such analysis deals with determining the risk level as well as main risk factors. In the specific case, an insurance company is faced with market share growth and profit decline. Discovered knowledge about the level of risk and main risk factors was not used to increase premium for the riskiest portfolio segments due to a specific market situation, which could lead to loss of clients in the long run. Instead, additional analysis was conducted using data mining methods resulting in a solution, which stopped further profit decline and lowered the risk level for the riskiest portfolio segments. The central role for the unexpected revealed knowledge in the chapter acts as the REFII model. The REFII model is an authorial mathematical model for time series data mining. The main purpose of that model is to automate time series analysis, through a unique transformation model of time series.
Chapter Preview


As data mining techniques have become more popular, they have become increasingly involved in risk evaluation in insurance companies (Apte, 1999; Chidanand, 1999; Pyle, 1999; Smith, 2000). Who is the riskiest client, who could have a car accident with highest probability during the lifetime of the contract, which is the low risk segment in insurance portfolio? This, and similar questions preoccupy portfolio managers in insurance companies.

Applied data mining model for the risk evaluation in insurance companies could vary (Apte, 1999). It can be built using probabilistic models (Chidanand, 1999), Fuzzy logic (Derik, 1995), Neural networks (Alexander, 1995), logistic regression (Kolyshkina, 2002), or linear models (Samson, 1997).

Complete Chapter List

Search this Book: