Trang 1 INSTITUTE OF ECONOMICS AND INTERNATIONAL BUSINESS---INTRODUCTION TO BUSINESS ANALYSIS:ANALYZING CUSTOMER BEHAVIOUR OF A CAR INSURANCE BUSINESSCourse code: VJPE205HK1-23241.1 Tra
FOREIGN TRADE UNIVERSITY INSTITUTE OF ECONOMICS AND INTERNATIONAL BUSINESS - INTRODUCTION TO BUSINESS ANALYSIS: ANALYZING CUSTOMER BEHAVIOUR OF A CAR INSURANCE BUSINESS Course code : Instructors : VJPE205(HK1-2324)1.1 Assoc.Prof Nguyen Thi Vinh PhD Pham Thi Cam Anh Group : 03 TEAM MEMBER Student’s name Student ID Nhữ Hà Phong 2112150133 Ngô Xuân Tuấn 2112150153 Vũ Thanh Đức 2112150054 Trần Nguyễn Tuấn Bách 2213530008 Hoàng Thuỳ Linh 2212550038 Cù Thảo Ly 2113150042 Hanoi, October 18th, 2023 Table of Content INTRODUCTION Chapter 01: DATA DICTIONARY Chapter 02: DATA CLEANSING & PREPROCESSING - DATA CLEANSING .6 - DATA PREPROCESSING Chapter 03: EXPLORING DATA ANALYSIS & FINDING BUSINESS INSIGHTS Part 1: Describing attributes: Creating real customer profile - Visualizing data of each attributes (features) - Five point summary of numerical attributes 23 - Creating real customer profile 23 Part 2: Multivariate statistics: Finding business insights 25 - Data visualization 25 - Business insights from data visuals 31 Part 3: Linear regression: Finding the trends and interdependence between outcome and other numerical features .33 - Methodology 33 - Model of theory .33 - Business insights from correlation matrix 34 Chapter 04: HYPOTHESIS TESTING .35 - Hypothesis testing for Part - Chapter 03: p-test 35 - Hypothesis testing for Part - Chapter 03: OLS method .40 Chapter 05: INSIGHTS, CONCLUSION & RECOMMENDATIONS 45 Part 01: Synthesis of business insights of real customer behavior 45 Part 02: Recommendations for the business 50 REFERENCES 51 APPENDIX: TEAM MEMBERS’ CONTRIBUTION ASSESSMENT 52 INTRODUCTION Globally, the car insurance sector stands out as the predominant segment within the property and casualty insurance industry, primarily owing to its substantial premium revenues Auto insurance holds significant importance as it provides coverage not only for physical damages resulting from accidents but also for any harm or injuries that may arise due to a vehicular accident, whether caused by the insured party or another vehicle or accident involving their own vehicle (Jennifer Rudden, 2023) The insurance status of a vehicle can provide insights into the characteristics of its driver (Stephanie Blows, 2003) Historically, car insurance premiums have been determined using self-reported information provided by the policyholder, with key factors including age, length of time with a driver's license, postal code, engine power, vehicle usage, and previous claims history (Roel Verbelen, 2018) This report should allow the insurance company to quantify the possibility of customers who can claim loans better and dig deeper into usage-based insurance Chapter 01: DATA DICTIONARY The dataset includes 18 features as described below: Table 1.1: Data dictionary No Field name Data type Description ID Integer Unique ID number for all customers Age Ordinal Range of all customers’ age (E.g: 26-30: aged from 26 to 30 years) Gender Nominal Gender of all customers Race Nominal Race of all customers Driving experience Ordinal Range of all customers’ total driving experience in year (E.g: 10-19y: having experience from 10 to 19 years) Education Nominal Education level of all customers Income Nominal Class of income of all customers Credit score Float Vehicle ownership Nominal Whether all customers own their vehicles ● 0: not owning the vehicle ● 1: owning the vehicle 10 Vehicle year Nominal The time that all customers have their vehicles 11 Married Nominal Whether all customers are married ● 0: not married ● 1: married 12 Children Nominal Whether all customers have children ● 0: not having children ● 1: having children 13 Postal code Nominal Postal code of where all customers live and Credit score of all customers use their vehicles Annual mileage of all customers using their vehicles 14 Annual mileage Integer 15 Vehicle type Nominal 16 Speeding violations Integer Number of speeding violations that all customers have committed while using their vehicles 17 Past accidents Integer Number of accidents that all customers have been involved in while using their vehicles Nominal Whether all customers claim their loans from the company ● 0: customers not claiming loan ● 1: customers claiming their loan 18 Outcome Type of all customers’ vehicle Chapter 02: DATA CLEANSING & PREPROCESSING - DATA CLEANSING 1.1 - Missing values INPUT: Check for missing values in the dataset =COUNTBLANK(A1:R2673) OUTPUT: The dataset does not contain any missing values, which is great The next step would typically be to look for any inconsistencies or outliers in the data 1.2 - Outliers a) INPUT: Find outliers in credit score column Q1=QUARTILE(H2:H2673,1) Q3=QUARTILE(H2:H2673,3) IQR=Q3- Q1 L Bound=Q1- IQR*1.5 U Bound=Q3+ IQR*1.5 Outliers = OR(H2$U$6) If false, it is not outlier; if true, it is outlier OUTPUT: Table 2.1: Finding outliers b) INPUT: Find outliers in annual mileage column Q1=QUARTILE(I2:I2673,1) Q3=QUARTILE(I2:I2673,3) IQR=Q3- Q1 L Bound=Q1- IQR*1.5 Document continues below Discover more from: International business Trường Đại học Ngo… 999+ documents Go to course 40 Samsung Electronic QUN TR HC International business 100% (25) BÁO CÁO CUỐI KỲ 29 CHUYÊN ĐỀ ĐỊNH… International business 98% (42) Pauline cullen the key 196 to ielts writing task International business 100% (19) Insurance International business 97% (38) Dinh huong chien luoc 32 kinh doanh quoc te c… International business 100% (16) Báo cáo thực tập 27 khóa - sv Nguyễn Min… International 100% (15) U Bound=Q3+ IQR*1.5 business Outliers= OR(I2$V$6) If false, it is not outlier; if true, it is outlier OUTPUT: Table 2.2: Finding outliers These outliers not have a statistically significant impact on the outcome of the analysis 1.3 – Duplicates INPUT: Remove duplicates Select all of the dataset(click on a random cell in the dataset then Ctrl+A) Data Remove Duplicates Select all Ok OUTPUT: No duplicate values found (Figure 2.1: Finding duplicate value) - DATA PREPROCESSING INPUT: Modify the dataset We assume that: Table 2.3: Modifying some features OUTPUT: Table 2.4: Output of modified dataset Chapter 03: EXPLORING DATA ANALYSIS & FINDING BUSINESS INSIGHTS Part 1: Describing attributes: Creating real customer profile - Visualizing data of each attributes (features) a - Age Row Labels Count Percentage 16-25 504 18.86% 26-39 857 32.07% 40-64 809 30.28% 65+ 502 18.79% Grand Total 2672 100% (Figure 3.1: Table and Pie chart presenting categories of age) Looking at the pie chart, we notice that the largest segment consists of individuals aged 26-39, accounting for 32% of the total The 40-64 age group is slightly lower at around 30% Meanwhile, both the 16-25 and over 65 age groups have roughly equal percentages, approximately 19% each From this, we can deduce that customers between the ages of 25-64 are more inclined to acquire insurance from the company