How deeply does Outlier Detection in Python cover label encoding?

Explains integer mapping and detector-specific behavior, but stops short of prescribing one universally best encoding.

Label encoding

As explained in Outlier Detection in Python

Ordinal encoding is a categorical encoding technique that maps discrete categories to integer codes representing an order.

Label encoding, also called ordinal encoding, replaces each categorical value with an integer code. For example, Color values Red, Blue, and Green can become 0, 1, and 2. The codes are labels, not measurements: their numeric order and spacing are arbitrary. section 4.9 section 9.4.2

The technique is commonly used when categorical features must be made numeric for an outlier detector. It avoids the extra binary columns produced by one-hot encoding, but the resulting representation must match the detector’s assumptions. section 3.3.2 section 9.4.2

Interpret label-encoded values according to the downstream detector:

Use ordinal encoding cautiously with distance-based detectors: arbitrary codes can create misleading distances between categories. section 9.4.2
Use it more favorably with Isolation Forest, which can treat repeated category values as duplicate records and exploit category frequency. section 9.4.2
Treat encoded categories as bins rather than ordered numeric measurements in HBOS; rare ordinal values contribute frequency information, not meaningful order. section 6.2

Department frequencies with Isolation Forest

Suppose a Department column contains 1,000 Sales records, 500 Marketing records, and 50 Engineering records, encoded as 0, 1, and 2. Isolation trees initially isolate category values similarly, then retain duplicate counts when nodes stop splitting. Engineering records require fewer estimated isolation steps and are therefore more likely to receive high outlier scores than common Sales records. section 9.4.2

How Outlier Detection in Python covers it

liveBook manning.com

Read Outlier Detection in Python on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

Machine Learning for Tabular Data: XGBoost, Deep Learning, and AIOrdinal encoding represents categorical values with integer codes that impose an ordered numerical representation.
liveBook manning.com

Grokking AI Algorithms, Second EditionLabel encoding maps each category to a numerical label, typically an integer in a fixed range.
liveBook manning.com

Ensemble Methods for Machine LearningOrdinal encoding maps each category of a categorical variable to a single numeric value, optionally preserving an inherent category order.

Compare how 4 books explain label encoding →

Related concepts

categorical encodingkind of · Categorical encoding transforms discrete category values into numerical representations, such as separate binary indicators or vectors with one active position per category.
one-hot encodingalternative to · One-hot encoding represents each category as a binary vector or set of indicators with exactly one active component identifying that category.
OrdinalEncoderused by · OrdinalEncoder is a scikit-learn transformer that maps categorical feature values to integer codes representing their ordered categories.
scikit-learnused by · scikit-learn is an open-source Python library providing tools for classical machine learning, including algorithms, preprocessing, model selection, and evaluation.
biasproduces · Bias is a systematic deviation or tendency in data, measurements, estimates, judgments, or outputs that skews them relative to a relevant target or impartial baseline.

This explainer was generated by Manning's knowledge graph from the text of Outlier Detection in Python. The section links point to the parts of the book it draws on.