Rare category

Q: How deeply does Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI cover rare category?

Defines rarity, explains its overfitting risk, and recommends aggregation while deferring implementation details to later methods.

As explained in Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI

A rare class is a target category represented by a small proportion of observations in a dataset.

A rare category is a categorical value that occurs infrequently or has very few instances in a dataset. Such categories can make a model overfit during training, so they should be identified early, ideally during data extraction. section 2.2.5

Handle rare categories by:

Aggregating similar rare categories into a larger class, using domain knowledge to decide which values belong together. section 2.2.5
Distinguishing rarity from high cardinality: too many labels may call for target encoding or embeddings rather than simple aggregation. section 2.2.5
Inspecting categorical distributions during exploratory analysis, alongside value counts and other checks for tabular-data problems. section 2.5

A practical encoding decision

Suppose a categorical feature contains several labels that appear only a few times. Combine genuinely similar labels into a broader category when domain knowledge supports that choice. If the feature instead contains many labels overall, consider target encoding for gradient boosting or embeddings for deep learning. section 2.2.5

How Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI covers it

Cover of Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI

liveBook manning.com

Read Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

Outlier Detection in PythonA minority class is a class containing substantially fewer observations than the majority class or classes in a classification dataset.
liveBook manning.com

Delivering Machine Learning Projects: From design to deploymentA rare entity is an infrequently occurring subject, case, or category represented in a data set.

Compare how 3 books explain rare category →

Related concepts

upsamplingconfigured by · Upsampling is a process that increases the number, density, or spatial or temporal resolution of samples or values in a dataset or representation by adding, duplicating, or estimating them.
fraudincludes · Fraud is intentional deception or unauthorized activity undertaken to obtain an improper benefit or cause another party a loss.
domain knowledgeused by · Domain knowledge is specialized understanding of the concepts, facts, practices, relationships, and constraints of a particular subject area.
anomalykind of · An anomaly is an observation, behavior, or state that deviates from an expected or typical pattern or condition.
rare wordcontrasts with · A rare word is a word that occurs substantially less often than other words in a reference text collection.

This explainer was generated by Manning's knowledge graph from the text of Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI. The section links point to the parts of the book it draws on.