ACME Insurance Inc. is an insurance company that offers affordable health insurance to thousands of customer all over the United States. The company sometimes find it difficult to estimate prices for new customers.
The company always use certain important criteria in estimating prices for new client by always referring back to similar previous historic data and thats quiet time wasting and the accuracies may be unusual. The goal of this project is to help the company way of achieving high accuracy in estimating the annual medical charges of their clients and to also save time.
I created automated system to estimate the annual medical expenditure for new customers, using information on same criteria or inputs they always use in estimation of their annual charges for thier customers.
The verified historical data of the company was available, consisting actual medical charges incurred by over 1300 customers. But it was difficult to understand the usage trends by age, sex, BMI, smoking habit and location across the regions. with the dataset in hand, I processed the data and performed Exploratory data analysis(EDA) on it.
after, I did the following to achieve my goal ;
- Explore the data and find correlations between inputs and targets
- Pick the right model, loss functions and optimizer for the problem at hand
- Scale numeric variables and one-hot encode categorical data
- Set aside a test set (using a fraction of the training set)
- Train the model
- Make predictions on the test set and compute the loss
Through the project, I learned a lot, from understanding the domain of the dataset to model creation in Jupyter notebook using python and it libraries.
I also worked to gather feedback on the project and made suggestions to the decision makers of the company to know the right things to work on #dataanalysis #machinelearning #datascience
Top comments (0)