BigML

Upload a data file or reference a remote URL to create a BigML Source — the starting point for every machine learning workflow.

Transform a BigML Source into a Dataset, applying field selections, filters, and sampling so the data is clean and scoped before model training begins.

Train a supervised decision-tree model on a BigML Dataset for classification or regression tasks.

Train a boosted or bagged ensemble of trees for higher-accuracy classification and regression — use when a single decision tree isn't accurate enough.

Run a single-row prediction against any trained BigML supervised model (decision tree, ensemble, logistic regression, or deepnet) and get the result back immediately.

Score an entire dataset at once against a supervised model and receive the results as a new dataset — the efficient path for bulk inference on thousands or millions of rows.

Group unlabelled data into k clusters using BigML's centroid-based clustering — use when you want to discover natural segments in your data without predefined labels.

Build an iforest-based anomaly detector on a dataset — the model identifies data points that deviate significantly from the norm, without needing labelled examples of anomalies.

Score a single data point against a trained BigML anomaly detector and receive a numeric anomaly score — scores near 1 indicate high anomaly likelihood.

Train a time-series forecasting model on sequentially indexed data — ideal for demand forecasting, sales projections, and trend analysis.

Generate a multi-step forecast from a trained BigML time-series model, returning predicted values for each future time horizon.

Train a logistic regression classifier on a BigML Dataset — a fast, interpretable model ideal for binary or multi-class classification with feature-weight transparency.

Poll a BigML resource (source, dataset, model, ensemble, cluster, anomaly detector, time series) for its current status — use to confirm training has completed before running predictions.

Run a WhizzML automation script in BigML — orchestrate multi-step ML workflows (data prep → train → evaluate → deploy) in a single API call.

Query BigML for a filtered, paginated list of any resource type (sources, datasets, models, ensembles, clusters, anomaly detectors) — use to inventory or audit your ML assets.