14.5: Key Terms

Key Terms and Definitions

• Ensemble Method — Combining multiple ML models to produce better predictions than any single model.

• Random Forest — An ensemble of bagged decision trees with random feature subsets at each split.

• Gradient Boosting — An ensemble method that builds trees sequentially, each correcting the previous one's errors.

• Deep Learning — Neural networks with multiple layers that learn hierarchical features from data.

• AutoML — Platforms that automate the ML workflow: preprocessing, model selection, hyperparameter tuning.

• Augmented Analytics — AI-assisted analytics that surfaces insights and assists human decision-makers automatically.

• SHAP — SHapley Additive exPlanations — a method assigning each feature a contribution score for a prediction.

• LIME — Local Interpretable Model-agnostic Explanations — approximating a complex model locally with a simple one.

• Explainable AI (XAI) — Methods and tools that make ML model decisions interpretable and transparent to humans.

• Data Mesh — A decentralized architecture where domain teams own their data as products with federated governance.

• Differential Privacy — Adding calibrated noise to query results so individual records cannot be identified.

• Federated Learning — Training ML models on decentralized data (e.g., on user devices) without collecting raw data centrally.

• Edge Analytics — Processing and analyzing data near the source (IoT devices, local servers) to reduce latency.

• MLOps — Practices for managing the ML lifecycle: deployment, monitoring, retraining, and versioning of models.

• AIOps — Using AI to manage IT operations — anomaly detection, predictive alerting, automated remediation.