Model Online / On-Site
Duration
Modules 5 modules

Course Objective & Scope

Used well, AutoML compresses weeks of trial-and-error into hours; used poorly, it hides data leakage and produces bloated ensembles that cannot be explained or deployed. This course covers open-source AutoML frameworks including AutoGluon, FLAML and Optuna-based pipelines, search-space and time-budget management, multi-layer stacking and model distillation.

The emphasis is on augmenting rather than replacing human expertise: enforcing the right validation split, feeding domain features into AutoML and choosing deployment presets that balance speed, accuracy and explainability. Participants produce both a strong benchmark and a slim production-ready model on an enterprise case.

Who Should Attend?

  • Data science teams building and maintaining many models
  • ML engineers who need rapid PoCs and strong benchmarks
  • BI and data analysts scaling their predictive capabilities
  • Prerequisite: Python and basic model evaluation knowledge

Key Takeaways

  • Build strong baselines on tabular and multimodal data with AutoGluon and FLAML
  • Enforce temporal and group-aware splits safely outside the AutoML black box
  • Constrain presets, pruning and model families to fit latency and compute budgets
  • Distil or refit heavy ensembles into compact models suitable for serving
  • Combine AutoML pipelines with domain feature engineering and business constraints

Curriculum

AutoML Architecture and Tool Ecosystem
  • The CASH problem: algorithm selection, HPO and ensembling
  • Comparing AutoGluon, FLAML, H2O and cloud AutoML offerings
AutoGluon and Multi-Layer Stacking
  • Automated handling of tabular, text and mixed columns
  • Repeated k-fold bagging and leak-free stacking layers
Resource-Aware Search with FLAML
  • Fast learner and hyperparameter search under tight compute budgets
  • Custom metrics, constraints and inference-latency limits
Pitfalls and Validation in AutoML
  • Target leakage, temporal splits and enforcing GroupKFold
  • How domain feature engineering multiplies AutoML performance
Production Readiness and Model Simplification
  • Ensemble pruning, refit_full and model distillation
  • Feature-importance reporting and packaged serving artefacts

Other Courses in This Category

  • Explainable AI (XAI)

    Glass-box models (EBM, GAM), TreeSHAP/KernelSHAP, ALE, counterfactual explanations (DiCE), uncertainty quantification and regulatory compliance for producing auditable, actionable explanations in high-stakes decisions.

    Machine Learning & Applied Data Science Corporate
    Live Course
    5 modules
    View
  • Causal Inference and Uplift Analytics

    Moving from correlation to causation with DAGs, propensity scores, Double Machine Learning (EconML/DoWhy), heterogeneous treatment effects (CATE) and Difference-in-Differences to measure price, campaign and policy impact.

    Machine Learning & Applied Data Science Corporate
    Live Course
    5 modules
    View
  • Federated Learning and Privacy-Preserving ML

    Collaborative model training across institutions or edge devices without centralising raw data: Flower framework, FedAvg/FedProx, non-IID data, differential privacy (Opacus) and secure aggregation.

    Machine Learning & Applied Data Science Corporate
    Live Course
    5 modules
    View
  • Introduction to Machine Learning - Regression Models Intro to Machine Learning - Regression Models

    Hands-on workshop covering Linear, Ridge, Lasso, and Polynomial regression models, feature engineering, MSE/RMSE/MAE/R² evaluation metrics, and regularization techniques.

    Machine Learning & Applied Data Science Individual Workshop
    Live Course
    3 modules
    View
  • Online Machine Learning and Streaming Analytics

    Incremental learning on streaming data with River, prequential evaluation, concept drift detection (ADWIN, Page-Hinkley) and contextual bandits for continuous adaptation in real-time systems.

    Machine Learning & Applied Data Science Corporate
    Live Course
    5 modules
    View
  • Graph Analytics and Graph Machine Learning

    Modelling relational data with NetworkX and PyTorch Geometric: centrality and community algorithms, graph features, Node2Vec and GCN/GraphSAGE/GAT for node, link and subgraph prediction in fraud rings and supply networks.

    Machine Learning & Applied Data Science Corporate
    Live Course
    5 modules
    View

Get in touch