國立虎尾科技大學 |

Probabilistic and Machine Learning Methods in Insurance.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Probabilistic and Machine Learning Methods in Insurance./
作者:	Lurvey, Carl M.
面頁冊數:	1 online resource (157 pages)
附註:	Source: Masters Abstracts International, Volume: 85-08.
Contained By:	Masters Abstracts International85-08.
標題:	Mathematics. -
電子資源:	click for full text (PQDT)
ISBN:	9798381721164

Probabilistic and Machine Learning Methods in Insurance.
Lurvey, Carl M.

Probabilistic and Machine Learning Methods in Insurance. - 1 online resource (157 pages)

Source: Masters Abstracts International, Volume: 85-08.

Thesis (M.S.)--California State University, Long Beach, 2023.

Includes bibliographical references

Insurance is one of the key industries supported by actuarial science. Whether it be life insurance, health insurance, or any other type of insurance, actuarial models are used in all manner of calculations. This thesis seeks to explore statistical and machine learning methodologies behind what is frequently referred to as "general insurance." This thesis will explore using statistical distributions to model claim severity and intensity. This thesis will also use machine learning techniques to model insurance claims.This thesis will then apply various statistical distributions to a real-world data set, the National Flood Insurance Program (NFIP), which is managed by the Federal Emergency Management Agency (FEMA). Through FEMA's open data portal OPEN FEMA, the public has access to decades' worth of data containing information on, among other things, policies, claims, and premiums. To model the frequency, this thesis will explore the Poisson and negative binomial distributions. To model severity, this thesis will explore Gamma, Pareto, and log-normal distributions. Parameters for the distributions will be estimated using both the Maximum Likelihood Estimation and the Method of Moments. The theoretical distributions will then be compared to the real-world data to determine which distributions provide the best fit. This thesis will also explore the modeling behind premium calculations. The frequency and severity distributions will be combined into compound processes to calculate expected claims over a specified time interval which will be used to drive premium calculations. As with the frequency and severity modeling, the NFIP data will be used to illustrate concepts in premium calculations. The NFIP data will also be used to compare how a strictly theoretical calculation of premium amounts compares to what premiums were actually charged.The statistical software R will be used to process the NFIP data and demonstrate the various concepts. R was chosen as the desired software due to the author's experience with it and its usage on many of the actuarial exams. To demonstrate various machine learning techniques, this thesis will then use a data set from Kaggle that contains information on automobile insurance claims. This thesis will explore a variety of methods including decision trees, random forests, gradient boosting support vector machine, and artificial neural networks. Just as was used with the NFIP data, R will be the software package used to analyze the data through the variety of machine learning techniques.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798381721164Subjects--Topical Terms:

527692
Mathematics.
Subjects--Index Terms:

Actuarial scienceIndex Terms--Genre/Form:

554714
Electronic books.

Probabilistic and Machine Learning Methods in Insurance.
LDR:03973ntm a22004097 4500 001 1152136
005 20241122094133.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798381721164
035 $a (MiAaPQ)AAI30691302
035 $a AAI30691302
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Lurvey, Carl M. $3 1479030
245 1 0 $a Probabilistic and Machine Learning Methods in Insurance.
264 0 $c 2023
300 $a 1 online resource (157 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Masters Abstracts International, Volume: 85-08.
500 $a Includes supplementary digital materials.
500 $a Advisor: Korosteleva, Olga.
502 $a Thesis (M.S.)--California State University, Long Beach, 2023.
504 $a Includes bibliographical references
520 $a Insurance is one of the key industries supported by actuarial science. Whether it be life insurance, health insurance, or any other type of insurance, actuarial models are used in all manner of calculations. This thesis seeks to explore statistical and machine learning methodologies behind what is frequently referred to as "general insurance." This thesis will explore using statistical distributions to model claim severity and intensity. This thesis will also use machine learning techniques to model insurance claims.This thesis will then apply various statistical distributions to a real-world data set, the National Flood Insurance Program (NFIP), which is managed by the Federal Emergency Management Agency (FEMA). Through FEMA's open data portal OPEN FEMA, the public has access to decades' worth of data containing information on, among other things, policies, claims, and premiums. To model the frequency, this thesis will explore the Poisson and negative binomial distributions. To model severity, this thesis will explore Gamma, Pareto, and log-normal distributions. Parameters for the distributions will be estimated using both the Maximum Likelihood Estimation and the Method of Moments. The theoretical distributions will then be compared to the real-world data to determine which distributions provide the best fit. This thesis will also explore the modeling behind premium calculations. The frequency and severity distributions will be combined into compound processes to calculate expected claims over a specified time interval which will be used to drive premium calculations. As with the frequency and severity modeling, the NFIP data will be used to illustrate concepts in premium calculations. The NFIP data will also be used to compare how a strictly theoretical calculation of premium amounts compares to what premiums were actually charged.The statistical software R will be used to process the NFIP data and demonstrate the various concepts. R was chosen as the desired software due to the author's experience with it and its usage on many of the actuarial exams. To demonstrate various machine learning techniques, this thesis will then use a data set from Kaggle that contains information on automobile insurance claims. This thesis will explore a variety of methods including decision trees, random forests, gradient boosting support vector machine, and artificial neural networks. Just as was used with the NFIP data, R will be the software package used to analyze the data through the variety of machine learning techniques.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Mathematics. $3 527692
650 4 $a Statistics. $3 556824
653 $a Actuarial science
653 $a Insurance claims
653 $a Machine learning
653 $a Statistical distributions
653 $a Neural networks
655 7 $a Electronic books. $2 local $3 554714
690 $a 0463
690 $a 0800
690 $a 0405
710 2 $a California State University, Long Beach. $b Mathematics and Statistics. $3 1182969
710 2 $a ProQuest Information and Learning Co. $3 1178819
773 0 $t Masters Abstracts International $g 85-08.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30691302 $z click for full text (PQDT)