liveBook · Manning

Random feature selection

As explained in Build AI Drug Discovery Pipelines

Feature subsampling is the random selection of a subset of input features for constructing an individual model component or tree split.

Random feature selection, also called feature subsampling, chooses only a random subset of input features when constructing each decision-tree split. In a Random Forest classifier, different trees therefore see different feature choices as well as different training-data samples. This decorrelates the trees, increases ensemble diversity, and helps reduce variance and overfitting while retaining the low bias of decision trees. section 5.3.2 section 5.5

A practical Random Forest workflow uses feature subsampling alongside ensemble evaluation:

  • Represent molecules as input features such as fingerprint bits. section 5.3.2
  • Train many decision trees, allowing each split to consider only a random feature subset. section 5.3.2
  • Combine the trees’ predictions to obtain a more robust classifier. section 5.3.2
  • Use out-of-bag samples to estimate performance without a separate validation set, especially when data are limited. section 5.3.2

CYP3A4 inhibition prediction

A Random Forest trained on molecular fingerprints achieved a validation PR AUC of 0.837, while its out-of-bag PR AUC was 0.806. Feature subsampling contributes to the ensemble’s diversity; feature importance can then rank influential fingerprint bits.

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains the mechanism and ensemble effects, then demonstrates evaluation, but does not detail the split-selection algorithm internally.
Examples
Uses molecular fingerprint bits, CYP3A4 inhibition classification, validation PR AUC, OOB PR AUC, and Gini feature importance.
What it assumes
Assumes basic decision-tree and classification vocabulary while defining feature subsampling and ensemble diversity directly.
What you can do after
Apply it via scikit-learn’s RandomForestClassifier and assess the resulting ensemble with validation or out-of-bag metrics.
How it connects
Connects feature subsampling with Random Forests, bagging, decision trees, overfitting, variance, and out-of-bag evaluation.

Also explained in

Compare how 2 books explain random feature selection →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.