A Computational Intelligence Approach for Predicting Medical Insurance Cost

In the domains of computational and applied mathematics, soft computing, fuzzy logic, and machine learning (ML) are well-known research areas. ML is one of the computational intelligence aspects that may address diverse difficulties in a wide range of applications and systems when it comes to exploitation of historical data. Predicting medical insurance costs using ML approaches is still a problem in the healthcare industry that requires investigation and improvement. Using a series of machine learning algorithms, this study provides a computational intelligence approach for predicting healthcare insurance costs. The proposed research approach uses Linear Regression, Support Vector Regression, Ridge Regressor, Stochastic Gradient Boosting, XGBoost, Decision Tree, Random Forest Regressor, Multiple Linear Regression, and k-Nearest Neighbors A medical insurance cost dataset is acquired from the KAGGLE repository for this purpose, and machine learning methods are used to show how different regression models can forecast insurance costs and to compare the models’ accuracy. The results shows that the Stochastic Gradient Boosting (SGB) model outperforms the others with a cross-validation value of 0.0.858 and RMSE value of 0.340 and gives 86% accuracy.

The risk of psoriasis in patients with uveitis: A nationwide population-based cohort study

Objective To evaluate whether the risk of subsequent psoriasis and psoriatic arthritis development is increased in patients with uveitis. Methods In Taiwan’s national health insurance research database, we identified 195,125 patients with new-onset uveitis between 2001 and 2013. We randomly selected 390,250 individuals without uveitis who were matched 2:1 to uveitis cases based on age, sex and year of enrolment. The characteristics of the two groups were compared. Using multivariate Cox regression, hazard ratios (HRs) for psoriasis or psoriatic arthritis corresponding to uveitis were computed after adjustment for age, sex, insurance cost and comorbidities. In subgroup analyses, separate HRs for mild psoriasis, severe psoriasis and psoriatic arthritis were calculated. Results The mean age of the study cohort was 50.2 ± 17.2 years. Hypertension, diabetes, hyperlipidaemia and obesity were more prevalent in the uveitis group (all p < 0.0001). The hazard of psoriasis or psoriatic arthritis development was significantly greater in the uveitis group than in the non-uveitis group (p < 0.0001); this increased risk persisted after adjustment for confounders [adjusted HR = 1.41; 95% confidence interval (CI), 1.33–1.48]. Adjusted HRs showed an increasing trend from mild psoriasis (1.35; 95% CI, 1.28–1.44) to severe psoriasis (1.59; 95% CI, 1.30–1.94) and psoriatic arthritis (1.97; 95% CI, 1.60–2.42). Conclusions This nationwide population-based cohort study revealed that patients with uveitis have an increased risk of subsequent psoriasis or psoriatic arthritis development.

A vastag- és végbéldaganat okozta országos epidemiológiai és egészségbiztosítási betegségteher Magyarországon

Összefoglaló. Bevezetés: Magyarországon a vastag- és a végbéldaganat mindkét nem esetében a harmadik leggyakoribb daganatos megbetegedés és a második leggyakoribb halálok. Célkitűzés: Elemzésünk célja volt a vastag- és végbéldaganat okozta éves epidemiológiai és egészségbiztosítási betegségteher meghatározása Magyarországon. Adatok és módszerek: Az adatok a Nemzeti Egészségbiztosítási Alapkezelő (NEAK) finanszírozási adatbázisából származnak, és a 2018. évet fedik le. A daganat típusait a Betegségek Nemzetközi Osztályozása (BNO, 10. revízió) szerinti C18-as, C19-es, C20-as, C21-es, D010–D014-es és D12-es kóddal azonosítottuk. Meghatároztuk az éves betegszámokat korcsoportos és nemek szerinti bontásban, a prevalenciát 100 000 lakosra, az éves egészségbiztosítási kiadásokat valamennyi ellátási formára és daganattípusra vonatkozóan. Eredmények: A vastag- és végbéldaganatok kezelésére a NEAK 21,7 milliárd Ft-ot (80,2 millió USD; 68,0 millió EUR) költött 2018-ban. A költségek 58,0%-át az aktívfekvőbeteg-szakellátás költségei teszik ki. Az összköltségek megoszlása szerint a legmagasabb költségek a férfiaknál (4,98 milliárd Ft) és a nőknél (3,25 milliárd Ft) is a 65–74 éves korcsoportban figyelhetők meg. A legnagyobb betegszámot a járóbeteg-szakellátás esetében találtuk: 88 134 fő, ezt a háziorvosi ellátás (55 324 fő) és a CT, MRI (28 426 fő) követte. A vastagbél rosszindulatú daganata esetében az egy betegre jutó aktívfekvőbeteg-kassza alapján az éves egészségbiztosítási kiadás 1,206 millió Ft (4463 USD/3782 EUR) volt a férfiak és 1,260 millió Ft (4661 USD/3950 EUR) a nők esetében. Következtetés: Hazánkban az aktívfekvőbeteg-szakellátás bizonyult a fő költségtényezőnek, mely magában foglalja az onkoterápiás gyógyszeres költségeket is. Orv Hetil. 2021; 162(Suppl 1): 14–21. Summary. Introduction: Colorectal cancer is the third most common type of cancer and the second most common cause of mortality in Hungary in both sexes. Objective: The aim of our study was to determine the annual epidemiological disease burden and health insurance cost of colorectal cancer in Hungary. Data and methods: Data were derived from the financial database of the National Health Insurance Fund Administration (NHIFA) of Hungary for the year 2018. Types of cancer were identified with the following codes of the International Classification of Diseases, 10th revision: C18, C19, C20, C21, D010–D014, D12. The data analysed included annual patient numbers according to age groups and sex, prevalence of care utilisation per 100 000 population, and annual health insurance costs for all types of care and all cancer types. Results: In 2018, NHIFA spent 21.7 billion HUF (80.2 million USD, 68.0 million EUR) on the treatment of colorectal cancer. 58.0% of the costs was spent on acute inpatient care. Regarding total costs, the highest costs were found in the 65–74 age group in both men (4.98 billion HUF) and women (3.25 billion HUF). The highest patient numbers were in outpatient care: 88 134 patients, general practice care (55 324 patients) and CT, MRI (28 426 patients). The annual health care treatment cost per patient was 1.206 million HUF (4463 USD/3782 EUR) in men and 1.260 million HUF (4661 USD/3950 EUR) in women. Conclusion: Acute inpatient care, including the costs of oncotherapeutic pharmaceuticals, was found to be the major cost driver in Hungary. Orv Hetil. 2021; 162(Suppl 1): 14–21.

Predict Health Insurance Cost by using Machine Learning and DNN Regression Models

Insurance is a policy that eliminates or decreases loss costs occurred by various risks. Various factors influence the cost of insurance. These considerations contribute to the insurance policy formulation. Machine learning (ML) for the insurance industry sector can make the wording of insurance policies more efficient. This study demonstrates how different models of regression can forecast insurance costs. And we will compare the results of models, for example, Multiple Linear Regression, Generalized Additive Model, Support Vector Machine, Random Forest Regressor, CART, XGBoost, k-Nearest Neighbors, Stochastic Gradient Boosting, and Deep Neural Network. This paper offers the best approach to the Stochastic Gradient Boosting model with an MAE value of 0.17448, RMSE value of 0.38018and R -squared value of 85.8295.

Climate Change and Insurance Business in Developing Economies

The capacity of insurance systems to reduce immediate losses caused by disasters through the provision of financial security against extreme weather conditions such as hurricanes, tropical cyclones, droughts, and floods avails an opportunity to developing countries for reduction of poverty and steady economic growth. The process of negotiating premium creates a platform for incentives to reduce risk and adapt to the climate change. There are opportunities for donors to combine resources so as to support the vulnerable communities with measures to reduce risk. In this research work, the scholar examines the process of financing disaster risk in developing countries, insurance of disaster in developing economy, advantages of disaster insurance, cost and risks involved in the insurance business, the concept of adaptation in insurance system, proposal of the Munich climate insurance initiative, the impact of insurance on the reduction of greenhouse gasses, and the globalisation of climate change risk.

Time-Varying Risk Aversion? Evidence from Near-Miss Accidents

We present evidence consistent with time-varying risk preferences among automobile drivers. Exploiting a unique dataset of agents’ high-frequency driving behavior collected by a mobile phone application, we show that drivers drive more conservatively following “near-miss” accidents. In a preferred specification, a nearmiss triggers a reduction in driving distance of 12.98 kilometers, in-car cellphone use by more than 100%, and highway use by 43.24%. Structural estimation results indicate that such changes in behavior are consistent with an increase in risk aversion of 10.54–43.77% and a reduction in annual insurance cost amounting to 2.04–3.31% of the average car insurance premium.