Concept · 1 paper · updated 2026-11-15
Target encoding
Replacing each level of a categorical feature with a number learned from the target, such as its mean outcome. Fitted on the rows it encodes, rare levels leak their own outcome. The study's best variant fits a random-intercept model and cross-fits the training rows. Author-reported: first for four of five learners in one lightly tuned benchmark, at much higher runtime with gradient boosting.
tabularcategorical-featurestarget-encodinghigh-cardinality
Markdown5 min read
AI-compiled overview, updated 2026-11-15, not yet reviewed by a human. Every claim cites its source; check it when a detail matters.
Overview
Target encoding turns each level of a categorical feature into a number learned from the outcome you want to predict. The simplest form is the mean outcome of the level in the training data 2104.00629 §1.2. The number then goes into the learner like any other numeric column.
The core risk is leakage. When a rare level's number is fitted on the same rows that get encoded, the level's number mostly reflects its own outcome. The model then trusts the feature too much and generalises poorly 2104.00629 §1.2. The two common fixes are regularisation, which pulls each level's number towards the overall average, and cross-fitting, which encodes each training row with a model that never saw that row 2104.00629 §1.2, §2.7.
Only one card in this wiki covers this slug, so every statement below comes from one benchmark study. For the wider family of encoders, see categorical encoding.
Variants
Impact encoding (smoothed target mean)
Each level is replaced by its conditional target mean, pulled towards the overall mean by a smoothing term. For classification the study uses one column per class, on a logit link 2104.00629 §2.6. The study sets the smoothing constant to 0.0001, which is almost no smoothing. The authors chose it so that this variant serves as a plain target-encoding baseline, and they call it the unregularised impact encoder in their results 2104.00629 §2.6, §4.3. A tuned smoothing value was not tested. The paper does not state how impact encoding handles levels unseen at prediction time 2104.00629 §2.6.
Without strong regularisation, impact encoding showed severe overfitting on the two datasets with the most levels 2104.00629 §4.2. Card: Pargent et al..
GLMM encoding with cross-fitting
A random-intercept (mixed) model is fitted per feature. Each level's encoded value is its conditional mode. Levels unseen at prediction time take the model's fixed intercept 2104.00629 §2.7. The authors describe this as a regularised version of impact encoding, with the amount of shrinkage set by the model rather than by a tuned constant 2104.00629 §2.7.
The study compares three settings:
- No cross-fitting (glmm-noCV). Training rows are encoded by the model that fitted them. It overfits like impact encoding on the high-cardinality datasets 2104.00629 §4.2.
- 5-fold cross-fitting (glmm-5CV). The headline variant. The prediction-time model uses all training rows, and each training row is encoded by the model that did not see its fold 2104.00629 §2.7.
- 10-fold cross-fitting (glmm-10CV). Often no better than 5-fold 2104.00629 §4.3.
For multiclass problems the study fits one one-vs-rest model per class, so each level gets one encoded column per class 2104.00629 §2.7. Card: Pargent et al..
When to use which
The numbers are the study's own and are author-reported on 24 OpenML datasets with five learners, lightly tuned 2104.00629 §3.3, §5.1.
| Situation | Use | Why |
|---|---|---|
| Accuracy matters most and a mixed-model fit per feature and fold is affordable | GLMM with 5-fold cross-fitting | Ranked first for every learner except SVM 2104.00629 §4.3. |
| No time to tune a smoothing constant on the whole pipeline | GLMM rather than impact | The authors say the shrinkage is set automatically 2104.00629 §2.7. This is their claim; the study does not test tuned impact. |
| Gradient boosting, tight runtime budget | Target-agnostic encoding such as frequency | Median runtime about 0.54 of one-hot, against 0.82 for impact and 7.21 for GLMM with 5-fold 2104.00629 tab. 1. |
| Cannot cross-fit | Do not use unregularised target encoding on high-cardinality features | The study found severe overfitting without regularisation or cross-fitting 2104.00629 §4.2. Use a target-agnostic encoder instead. |
| Impact encoding with a tuned smoothing value | Untested here | The study fixes the smoothing at 0.0001 2104.00629 §2.6. Check the result yourself before relying on it. |
| Many classes | Budget for one model per class | GLMM fits one one-vs-rest model per class 2104.00629 §2.7. |
| Support vector machine | Check the figures before choosing | GLMM with cross-fitting did not rank first for SVM 2104.00629 §4.3. |
Runtime grows with cross-fitting. Median runtime relative to one-hot for gradient boosting rose from 1.65 with no cross-fitting to 7.21 with 5 folds and 12.71 with 10 folds 2104.00629 tab. 1. Under the study's setup this is the main cost of the best variant.
The study tuned the learners only lightly and ran 5-fold CV once. It also used 24 datasets, too few for the authors to link the ranking to dataset traits 2104.00629 §5.1, §5. Do not assume the ranking transfers to tuned pipelines or to your data.
Implementation sketch
A generic cross-fitting wrapper, then a fitter for impact encoding. The wrapper is the one the study uses with GLMM. Using it with impact encoding is an extension the study did not test 2104.00629 §2.6, §2.7.
# Cross-fitting wrapper for any target-based level encoder [arxiv:2104.00629 §2.7]
# Input: levels (n,) categorical; target (n,)
# fit_values(levels, target) -> (values per level, fallback for unseen)
# K folds; K = 1 means no cross-fitting
def encode_train(levels, target, K, fit_values):
# Encoder used at prediction time, fitted on all training rows [§2.7]
full = fit_values(levels, target)
if K == 1:
# Rows are encoded by the model that saw them: prone to leakage [§4.2]
return lookup(full, levels), full
encoded = empty(n)
# Out of fold: each row is encoded by a model that never saw its fold [§2.7]
for fold, rest in make_folds(n, K):
values = fit_values(levels[rest], target[rest])
encoded[fold] = lookup(values, levels[fold])
return encoded, full# Impact fitter: per-level mean shrunk towards the global mean [arxiv:2104.00629 §2.6]
# Regression shown; classification uses one column per class on a logit link
def fit_impact(levels, target, eps):
# Overall mean, which the shrinkage pulls towards [§2.6]
global_mean = mean(target)
values = {}
for level in unique(levels):
rows = levels == level
# shrink() is not given as a formula in the paper; eps = 0.0001 in the study [§2.6]
values[level] = shrink(mean(target[rows]), count(rows), global_mean, eps)
# Assumption: the paper does not state a fallback for unseen levels with impact [§2.6]
return values, global_meanAt prediction time, encode new rows with the full-data model and its fallback. The ranking depends on how unseen levels are handled, and the paper does not state that rule for impact encoding, so check it before comparing numbers 2104.00629 §2.6, §2.7.
Contradictions
None known within this wiki. The study has one source, so no two papers disagree. The released code differs from the paper in several places. It does not include the impact encoder that its benchmark calls, so impact results cannot be checked against code. The card lists the other differences, such as how the high-cardinality threshold is counted. Those may change numbers reproduced from the code 2104.00629 §2.6. See Pargent et al. for details.
Papers
- Pargent et al.: compares regularised target encoding with cross-fitting against six other encoders on 24 datasets with five learners. It finds the GLMM variant with 5-fold cross-fitting ranked first for every learner except SVM (author-reported).
Source and license
Compiled from the papers listed on this page, all published under CC BY, CC BY-SA or CC0, with citations. See each paper's page for its authors and license.