About

This canvas is focused on analyzing and predicting printer maintenance needs using synthetic data. It generates a dataset of printer health updates with 20 features and a binary target indicating whether maintenance will be needed. Then, it performs feature importance analysis using a Random Forest model to identify key predictors. Additionally, it builds and evaluates a logistic regression model to predict maintenance needs, including producing classification metrics and ROC curves. Finally, it assesses the cost optimization of proactive versus reactive maintenance strategies by calculating total costs across different risk thresholds to find the optimal decision point.