Knowledge Graph — Coursera Notes › Academic disciplines › Information Technology / Computer Science › Data management

Data quality issues

concept · part of Data management

Data quality issues arise when incoming data has missing values, outliers, or shifts in distributions, degrading model performance. Causes include sensor malfunctions, manual data entry errors, changes in data collection methods, or inconsistent data formats. Mitigation involves setting up data validation pipelines that automatically check for consistency, completeness, and adherence to expected ranges. Tools such as Azure Data Factory or Great Expectations can preprocess data and handle missing values or outliers before feeding to the model. For example, a healthcare company predicting patient outcomes must ensure data from medical devices is consistent and accurate via validation pipelines.

This is the text view of an interactive 3D knowledge graph — open this page with JavaScript enabled to explore it visually.