liveBook · Manning

SGDRegressor

As explained in Build AI Drug Discovery Pipelines

SGDRegressor is a Scikit-Learn estimator that fits linear models by minimizing a regularized empirical loss with stochastic gradient descent.

SGDRegressor fits regularized linear models with stochastic gradient descent. With epsilon-insensitive loss, it provides a support vector regression (SVR) approach that ignores small errors and can resist noisy solubility labels. section 4.3.3

A practical molecular-solubility pipeline expands engineered descriptors with degree-two polynomial features, imputes missing values, standardizes them, and enables L2 regularization, inverse-scaling learning, and early stopping. The selected model reached validation RMSE 1.3791 and test RMSE 1.2783, beating ESOL and a y-scrambled baseline. section 4.3.3

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains loss choices, learning-rate schedules, regularization, and early stopping, but defers broader comparisons to an appendix.
Examples
Demonstrates molecular descriptors, degree-two features, loss curves, and solubility RMSE comparisons.
What it assumes
Assumes engineered molecular descriptors and separated training, validation, and test data.
What you can do after
Apply it via Scikit-Learn to train and evaluate a regularized descriptor-based regression pipeline.
How it connects
Connects SVR with Ridge regression, Lasso, RANSAC, Huber regression, polynomial features, and ESOL baselines.

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.