liveBook · Manning

Epsilon-insensitive loss

As explained in Build AI Drug Discovery Pipelines

Epsilon-insensitive loss assigns zero penalty to errors within an epsilon margin around the target and increases linearly for errors outside that margin.

Epsilon-insensitive loss ignores small prediction errors inside an epsilon margin and penalizes larger errors linearly. This makes regression less sensitive to noisy experimental solubility labels. section 4.3.3

Use it when:

  • A descriptor-based regression problem contains noisy measurements or outliers. section 4.3.3
  • A linear model operates in an expanded feature space, such as polynomial molecular descriptors. section 4.3.3
  • Small errors are practically unimportant and should not dominate optimization. section 4.3.3

Solubility regression pipeline

An SGDRegressor uses loss="epsilon_insensitive" with L2 regularization, early stopping, inverse-scaling learning rate, imputation, polynomial features, and standardization. The resulting model achieved a test RMSE of 1.2783, below ESOL's 1.3787 baseline. section 4.3.3

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains the error-insensitivity and outlier rationale, then demonstrates configuration without deriving the loss mathematically.
Examples
Uses molecular descriptors, polynomial features, SGDRegressor, and ESOL and y-scrambled baselines for solubility prediction.
What it assumes
Assumes familiarity with regression models, engineered molecular descriptors, training data, and validation or test RMSE.
What you can do after
Apply epsilon-insensitive loss through Scikit-Learn's SGDRegressor and compare resulting regression performance against baselines.
How it connects
Connects support vector regression with Ridge regression, Lasso, RANSAC, Huber loss, squared-error loss, and early stopping.

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.