AutoAnomaly

Use DataRobot Autopilot to train diverse anomaly detections models on a dataset.

Prediction and deployment methods execute on the anomaly detection model with the highest synthetic AUC at the time of calling. Training is performed within a new, automatically created DataRobot project.

Usage

Train

import pandas as pd
import datarobotx as drx

train = pd.read_csv('https://s3.amazonaws.com/datarobot_public_datasets/anomaly_use_cases/network_intrusion/Anomaly_Detection_kdd99_train_nolabel_v2.csv')
model = drx.AutoAnomalyModel().fit(train)

Predict

test = pd.read_csv('https://s3.amazonaws.com/datarobot_public_datasets/anomaly_use_cases/network_intrusion/Anomaly_Detection_kdd99_test.csv')
predictions = model.predict(test)
anomaly_scores = model.predict_proba(test)

Deploy

deployment = model.deploy()

Time Series

Time series anomaly detection models are also supported. See the AutoTS docs for an example.

API Reference

AutoAnomalyModel([name])	Automated anomaly detection orchestrator.
AutoTSModel([name, feature_window, ...])	AutoTS orchestrator.
DRConfig([Data, Target, Featurization, ...])	DataRobot configuration.
Deployment([deployment_id])	DataRobot ML Ops deployment.