[SQL-Python-ML]
Please notice that most of the visuals I plotted are Plotly Express and many platforms have not supported this Plotly libraries, If you cannot see visualizations from the ipynb file, please look for it in the PDF file. I am sorry for this disadvantageous, please sympathize.
In 2023, Sunrise Life Insurance is looking to anticipate which customers are likely to lapse on their policies. By identifying such customers, the company intends to proactively reach out to them with some customer care program in a bid to retain them.
To achieve this, Sunrise Life Insurance will need to employ predictive analytics techniques that can analyze vast amounts of data on customers and other relevant factors. Once the predictive model has been developed, the company will use it to generate a list of customers who are at high risk of lapsing. The company will then reach out to these customers with tailored offers that are designed to meet their unique needs and preferences.
The goal is to leverage data insights to improve customer satisfaction, reduce lapse, and increase customer loyalty.
- SQL - Provide needed data for The Marketing department
- Data Wrangling
- Define types of Missing Data (MCAR, MAR, MNAR)
- Use MICE imputation for missingness (in Annual Income)
- EDA (visualize to find insights), Data visualization with Pygwalker
- Statistical Analysis (Hypothesis Testing with Mann-Whitney-U)
- Building a classification model to predict a customer who is likely to lapse based on the provided dataset.
- Label Encoder
- Upsampling data (imbalanced data)
- Features selection (task required)
- Scale data (Min-max scaler)
- Conduct marketing strategies (Chiến dịch heart-touching, thay đổi nhận thức khách hàng)
- Focus on exploiting the elderly group (Khai thác nhóm lớn tuổi, đặc biệt là phụ nữ)
- Strengthen consultancy activities (Tăng cường công tác tư vấn trung thực, trách nhiệm)
- Expand distribution channels (Khai thác kênh phân phối tại các Ngân hàng qua kênh Bancassurance)
- In the case of low-income pensioners (Nghiên cứu mở rộng sản phẩm cho đối tượng thu nhập khác)
- Promote the development of the life insurance industry in Vietnam, aiming to gain 15% of population using life insurance in 2030 (According to the Vietnam Insurance Association, this rate has only reached less than 10%).
- Create trust for owners, improve 20% revenue in the following years.