Hi. This sounds similar to a project I did for a Health Insurance company building a Health Score based on historical patient data. I will need more details from you, but I'm assuming this is a standard prediction model of estimating the probability of diabetes given some clinical or registration data. I'm also assuming this to be the level of complexity of a prototype, not a production-ready model. So the deliverable should be:
* A Jupyter notebook with Python code for the data processing pipeline, machine learning model + parameters, evaluation metrics, and descriptive analysis of the results.
This project can be delivered in 2-3 days if these assumptions hold:
* Dataset is in English.
* Dataset is non-normalized.
* Single target variable with probability of a binary outcome.
* No minimum requirement of model efficiency.
* No need for a test suite, deployment or setup scripts.
More complexity can be added to the project if necessary, including the need for production-ready code, but it will have to be priced accordingly.
Thank you and I am looking forward to working with you.