The library goes far beyond basic ML. These are one-liners for problems that usually need specialist packages.
causal.estimate_ate(df, treatment, outcome) 路 causal.uplift(...)Not just correlation, causation: estimate the average treatment effect of an intervention, and model who responds to it (uplift).
In plain English: Goes past 'these move together' to 'this actually caused that', and finds who responds to a change.
import breezeml
breezeml.causal.estimate_ate(df, treatment="promo", outcome="spend")conformal.conformal_regressor(model, df, target, alpha=0.1)Prediction intervals with guaranteed coverage: 'the price is 40k to 55k, 90% of the time'.
In plain English: For number predictions: a guaranteed range, like 'the price lands between 40k and 55k about 90% of the time'.
import breezeml
lo, hi = breezeml.conformal.conformal_regressor(model, calib, "price", alpha=0.1)anomaly.compare(df) 路 isolation_forest 路 local_outlier_factor 路 one_class_svmFind the weird rows with four methods, or compare them all at once.
In plain English: Finds the odd rows that do not fit the rest, with several methods you can compare.
import breezeml
breezeml.anomaly.isolation_forest(df, contamination=0.05)timeseries.make_features(...) 路 timeseries.forecast(...) 路 timeseries.compare(...)Build lag/rolling features, compare forecasting models with proper time-aware splits, and forecast ahead.
In plain English: For data over time. It builds date-aware features and forecasts ahead without peeking at the future.
import breezeml
feat = breezeml.timeseries.make_features(df, date_col="date", target="sales")
breezeml.timeseries.forecast(df, date_col="date", target="sales", horizon=30)survival.kaplan_meier(df, duration_col, event_col)Time-to-event modeling: how long until churn, failure, or recovery, with censoring handled correctly.
In plain English: Models how long until something happens, like churn or failure, even when some cases have not happened yet.
import breezeml
breezeml.survival.kaplan_meier(df, "tenure_months", "churned")multi.multi_label(df, targets) 路 recommend.collaborative_filter(...)Predict several labels at once, or build a collaborative-filtering recommender from a ratings table.
In plain English: Predict several labels at once, or recommend items based on who liked what.
import breezeml
breezeml.multi.multi_label(df, targets=["tag_a", "tag_b"])active.query(...) 路 semisupervised.self_train(...) 路 autofeat.engineer(...)Label smartly (active learning), learn from mostly-unlabeled data (self-training), and auto-engineer new features.
In plain English: Spend labeling effort wisely, learn from mostly-unlabeled data, and auto-invent useful new features.
import breezeml
next_to_label = breezeml.active.query(model, unlabeled_df, n=10)
breezeml.autofeat.engineer(df, "target")text.embed(df, text_columns)Turn free-text columns into numeric embeddings you can feed to any model (needs the sentence-transformers extra).
In plain English: Turns free-text columns into numbers a model can actually use.
import breezeml
df2 = breezeml.text.embed(df, text_columns="review")