پژوهشنامه بیمه
سال سی و نهم شماره 1 (زمستان 1402)

The healthcare insurance industry faces a significant challenge predicting individuals' insurance costs, which are based on complex parameters such as age and physical characteristics. Insurance companies categorize policyholders into high-risk and low-risk groups to manage risks and avoid potential losses. However, the accurate estimation of costs for each individual can be a daunting task. By leveraging data science and machine learning techniques, insurance companies can improve their cost estimation accuracy and better manage risks. This approach can help insurance companies to provide more accurate insurance coverage and pricing for individuals leading to higher customer satisfaction and lower financial losses. To address this challenge, a data science and machine learning-based approach that uses ensemble learning to predict high-risk and low-risk individuals is used. The method involves several steps including data preprocessing, feature engineering, and cross-validation to evaluate the model's performance. The first step involves preprocessing the data by cleaning it, handling missing values, and encoding categorical variables. The second step generates new features using feature engineering techniques such as scaling, normalization, and dimensionality reduction. Next, ensemble learning is used to combine multiple regression methods such as logistic regression, neural networks, support vector machines, random forests, LightGBM, and XGBoost. By combining these methods, the aim is to leverage their strengths and minimize their weaknesses to achieve better prediction accuracy. Finally, the model's performance is evaluated using cross-validation techniques such as k-fold cross-validation. These techniques help to validate the model's accuracy and prevent overfitting. The proposed approach achieves an AUC of 0.73 demonstrating its effectiveness in predicting high-risk and low-risk individuals. In conclusion, the healthcare insurance industry can benefit greatly from data science and machine learning-based approaches. By accurately predicting high-risk and low-risk individuals, insurance companies can better manage risks and provide more accurate coverage and pricing for their customers. This can lead to the improvement of  customer satisfaction and the reduction of financial losses for insurance companies.

The accurate and scientific assessment of the risk to issue an insurance policy is one of the most critical and important stages of risk assessment frameworks. This leads companies to identify high-risk customers and determine the policy rates in accordance with their risks, and as a result, the claims will be covered appropriately through the insurance premiums. In this paper, a new method is presented to define the concept of risk factor in more practical, flexible and accurate way. In this method, which is based on an unsupervised clustering algorithm, initially, every single factor is examined based on different ranges and their corresponding impact on customer loss levels. Then, considering their connection with the ranges of other factors in terms of creating similar levels of customer loss, they are combined to form a package. Thus, different packages are created, each of which is considered a risk factor and comprise the ranges of factors affecting different levels of loss. The k-means clustering method was used to divide insurers into clusters with similar risks, which correspond to the risk packages associated with the customers' risk level. The number of desired clusters should be determined in advance, which is the main challenge of using this algorithm. Two main approaches for validation, namely the silhouette score and the elbow method, were presented. Based on the elbow plot and silhouette coefficient, as well as considering the practical and realistic evaluation needed by insurance companies, four clusters were obtained. Cluster 2 and 3 are similar and can be merged to form a cluster of medium risk level. Therefore, three clusters were considered the best outcome for categorizing insurance policyholders. The risk packages can be introduced from the examination of the 3 clusters including People with high, medium and low age (confidence interval) with low price car whose gender is male can be introduced as the highest level of risk; People with medium and high ages (confidence interval) with medium and high car prices can be considered as medium risks, and Middle-aged and older people (confidence interval) with expensive cars were considered the lowest level of risk. From the results of these risk packages, it can be concluded that although a significant population of older policyholders falls into the first package (first cluster), they have the highest level of risk. On the other hand, the older people in the third package (even though their average age is the highest among the clusters) have the lowest level of risk. Another important point is that the risk level decreases as income increases simultaneously with age.

The use of new marketing techniques in today's businesses is highly needed by all organizations. One of the important issues in the field of customer retention is the customer's lifetime value. Customer lifetime value has a great impact on optimizing the performance of companies, including insurance companies. The purpose of this research is to identify the influencing factors on increasing the customer lifetime value of insurance companies. This research is a type of mixed research with a qualitative and quantitative approach, which is a survey study in terms of its purpose, application, and in terms of data collection. The research population includes managers, experts, and university professors from insurance companies and organizations. who were selected using the snowball sampling method. In the qualitative part, the data collection tool was an interview, and in the quantitative part, a questionnaire was used to identify the categories, and a semi-structured interview was used, and a questionnaire was used to validate the model. In the qualitative part of the data analysis method, the data theory approach was based on the Strauss and Corbin method, which was compiled using MAXQDA software and using the coding method, and in the quantitative part, the analysis method was based on Kendall's correlation test. In order to examine the validity of the research in the qualitative part, the Cressol model was used along with content validity and intra-coder and inter-coder reliability, and in the quantitative part, in order to test the validity of the research, content validity and retest validity were used. The findings of the study identify 9 causal factors, 3 strategic factors, 4 intervening factors, and 3 contextual factors that contribute to increasing customer lifetime value in the insurance industry. The conclusion of the research is the presentation of a model that includes causal, contextual, and intervening conditions, along with strategies to increase customer lifetime value and their consequences in the insurance industry.

Weibull distribution, introduced by a Swedish physicist named Weibull, is the most common model used in studies of reliability, longevity, quality control. It is widely used in various fields of science including insurance, medicine, and engineering . This distribution is flexible enough to model different data. The main goal of this research is to calculate insurance premiums and estimate Weibull distribution parameters using various estimation methods. In this article, the parameters of Weibull distribution and net premium have been estimated using moment estimation, maximum likelihood, least squares of error, weighted least squares, percentage, Cramer-Von- Mises, mixture of moment and maximum-likelihood against outliers. R software was used for simulation and numerical calculations purposes and also Easyfit software was used to fit Weibull distribution to the real example data. In the end, two real data examples for obtaining various estimators of the premium in case of unknown parameters β and θ and known α are presented. In this research bias, the mean square error of net premium and unknown parameters β and θ were obtained using different estimators for Weibull distribution data as well as the generalized variance of unknown parameters β and θ. In this part, the evaluation and comparison of the estimators using real and simulated data was done, which was obtained by different for real data. For example, in the moment method, was equal to 5, based on which the net premium is 3.37657. In the simulated data, according to k (number of outliers), n (sample size) and β and θ values, bias values, mean squared error and generalized variance of premium and different estimators were obtained. As an example, for n=10, k=1, β=1.5, θ=3 and α=70, by comparing the bias and generalized variance of the estimators, we come to the conclusion that based on the bias, the percentile estimator has a better performance than the other estimators. Simply put, it has less bias and according to the generalized variance, the maximum likelihood estimator has a better performance than other estimators and the estimators are consistent (the generalized variance decreases with the increase of the sample size). Based on the mean square error, the moment estimation has a better performance than other estimators.

Profitability is a crucial indicator of insurance company performance, as it reflects their ability to invest and grow. Supervisors also rely on financial characteristics, including profitability, to determine the viability of insurers. The data envelopment analysis (DEA) method has been widely used to evaluate the performance of insurance companies. However, conventional financial ratios are often lacking in such studies, making DEA an effective alternative for measuring profitability. Given the rapid growth of the insurance industry in Iran, this research aims to examine the profitability of non-life insurance companies in Iran and expand coverage analysis. The study focuses on 18 insurance companies listed on the Tehran Stock Exchange from 2013 to 2014, with complete and available information during the research period. Profitability is measured using DEA, and the Tobit estimator model is used to investigate the impact of company size, company age, and product variety on profitability. The results highlight the importance of properly managing expenses and incomes for insurers. Additionally, the study finds that company size, company age, and product variety do not have a significant relationship with profitability. Insurance companies need to effectively organize their expenses and various types of business to maximize profit ratios. Optimizing cost structures and diversifying business ventures are key strategies for achieving profitability. This research provides insights into insurance company performance, allowing for a better understanding of their relative profitability within the industry and the formulation of targeted strategies. It also aids the government in developing policies that support the industry's growth.

Data mining is known as a process of discovering patterns in large datasets through a combination of statistical tools and techniques. In recent years, data mining and its applications in different businesses have increasingly grown. Insurance industry is one of the data-driven businesses whose survival is so dependent on satisfying customers besides achieving the highest benefit. Information or data is a vital asset of the insurance industry ;accordingly, using data mining techniques to discover patterns behind large datasets is a need. Having seen the increasingly high rate of information technology and recorded data in data-driven businesses, lots of industries like the insurance industry have been urged to use state-of-the-art data mining techniques to turn raw data into useful information using Big Data Analytics. Looking at the current research on data mining applications in the insurance industry proves the fact that we should recognize the state-of-the art techniques in data mining and set new strategies to focus on Big Data Analytics more. Big Data Analytics consists of the algorithms which are more efficient and less time-consuming so it can help to identify patterns and rules in complex datasets. For this purpose, this paper presents a comprehensive literature review regarding the usage of data mining techniques in the insurance industry by the scientometrics approach. For this purpose, first we searched and gathered bibliometrics files of recent researches from Web of Science and Scopus into four different scenarios. In each scenario, we looked up for different keywords regarding “Data Mining”, “Insurance Industry”, and “Risk Management” to make sure that all the results would be specifically focused on the research topic. Then, we used R programming software to analyze the results of each scenario based on keywords co-occurrence in the given research. The results of keywords co-occurrence and a word cloud of recent research confirm that insurance companies should focus on Big Data Analytics instead of traditional data processing to get information systematically from too large or complex datasets. Big Data Analytics has been used for several years, but in recent years many data-driven businesses, like the insurance industry, have used its techniques associated with risk and risk factor identification. Risk management in the insurance industry has been widely considered in recent researches. Therefore, in this paper, some high-ranked journals and the most significant researches have been identified and recommended in order to pave the way for future researches in this field. We hope that the comprehensive literature review provided in this paper can help the researchers to focus on the relative journals and researches published then get into more details. For this purpose, the lists of all journals and conferences besides the most cited researches are provided in the experimental section of this paper. Also, the ranking list of different countries from all around the world related to data mining and Big Data Analytics in the insurance industry is presented. The results show that Iran is the 15th  country that uses data mining techniques and it is the 17th  country in the world focusing on risk management in the insurance industry.