Imputation
Imputation is the replacement of missing or incomplete data with estimated, inferred, representative, or otherwise selected values.
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
Imputation is the replacement of missing data with estimated values or classes derived from available information.
- How it is used
- Prepares tabular datasets for algorithms that lack native missing-value support while preserving informative missingness patterns.
- How deep it goes
- Defines univariate and multivariate replacement, compares their errors, and defers specialized treatment of MNAR data.
- Examples
- Compares mean-based SimpleImputer with Random Forest-based IterativeImputer on continuous tabular features and mean absolute error.
- What you can do after
- Apply imputation with scikit-learn’s SimpleImputer or IterativeImputer and evaluate replacement quality with mean absolute error.
Read how Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI explains imputation →
Financial AI in Practice
The replacement of missing data with estimated or designated values.
- How it is used
- Supports credit-risk data preparation by addressing incomplete application fields that could bias models or distort risk assessment.
- How deep it goes
- Compares numeric estimates, missing bins, and row deletion, but gives no imputation procedure.
- Examples
- Uses omitted employment status as a possible default-risk signal and numeric fields as imputation targets.
- What you can do after
- Evaluate whether to exclude incomplete rows, estimate numeric values, or retain missingness as a separate risk signal.
Also covered in
Related concepts
- missing valueused by · A missing value is an expected data value or observation that is absent, unavailable, unknown, or inapplicable.
- multivariate imputation by chained equationincludes · Multivariate imputation by chained equation is a multiple-imputation technique that iteratively models each incomplete variable using other variables to estimate missing values.
- scikit-learnused by · scikit-learn is an open-source Python machine-learning library providing algorithms and tools for data preprocessing, model fitting, selection, and evaluation.
Questions about imputation
What is imputation?
Imputation is the replacement of missing or incomplete data with estimated, inferred, representative, or otherwise selected values.
Which Manning books explain imputation?
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI and Financial AI in Practice.
What is the difference between imputation and omission?
Imputation is the replacement of missing or incomplete data with estimated, inferred, representative, or otherwise selected values. Omission is the handling of missing data by excluding incomplete records or fields.
What kinds of imputation are there?
Kinds of imputation include multivariate imputation by chained equation, mean imputation and univariate imputation.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

