Skip to main content

Skills application

Instruction and application
In Progress

The data checkup lab​

In this skills application, you will prepare a high-quality training dataset for a hospital readmission model by splitting data correctly, addressing class imbalance, and resolving data quality issues.

<g></g><defs><clipPath><rect width="24" height="24" fill="white"></rect></clipPath></defs>## Context Organization: You have joined MediCareAI, a healthcare analytics company developing machine learning models to predict patient readmissions and support hospital resource planning.

Current challenge: The team’s dataset contains patient demographics, clinical records, and admission histories, but it suffers from imbalance, missing values, and potential data leakage. These issues are causing unreliable model evaluation and inconsistent performance across validation runs.

Your role: As the data scientist on the project, your task is to prepare the dataset for model training by designing robust data splits, addressing class imbalance, and resolving key data quality issues to ensure fair and reliable learning.

Objective: By the end of this activity, you will have created a well-prepared training and validation dataset that minimises leakage, improves model generalisation, and enhances prediction reliability.

Success criteria​

To complete this skills application successfully, you should demonstrate:

  • Correct application of a suitable data splitting strategy that prevents leakage.
  • Effective handling of class imbalance using appropriate resampling or weighting methods.
  • Implementation of data quality checks and corrections for missing values and outliers.
  • Clear documentation or code comments explaining each decision and how it improves model reliability.

Instructions and materials​

Follow the steps below to complete this skills application. Completing this activity will “unlock” the solution example on the following page.

Download the activity resources​

Select the links below to download the Jupyter Notebook anddataset for this activity. The notebook guides you through each step of the process, and the dataset provides the data you’ll need to perform the data preparation and model reliability tasks.

Skillable_lab_M7_Unit3_Data_Checkup_Lab.ipynb

MediCareAI_readmissions.csv

Access the Jupyter Notebook and dataset​

  • Open the provided Jupyter Notebook for this activity and download the accompanying dataset.
  • Upload the dataset to your notebook environment before running the code.
  • The notebook contains all baseline code and step-by-step guidance to help you complete the data preparation and model reliability tasks.

Complete the tasks​

  • Design and apply a data splitting strategy that prevents leakage and supports reliable model evaluation.
  • Implement techniques to handle class imbalance and improve model fairness and performance.
  • Develop and apply data quality protocols to clean, validate, and prepare a trustworthy training dataset.

Record and submit​

Document your findings and reflections directly in the notebook. Save your completed file and submit it for review.