Course Objective & Scope
In system logs, sensor streams, transaction records or business metrics, anomalies are rare and labels are usually missing. The programme compares methods from z-scores and robust statistics to Isolation Forest, Local Outlier Factor and One-Class SVM, autoencoder-based approaches and time series anomaly detection, using libraries such as PyOD.
Evaluation with few labels, threshold tuning, alert prioritisation and an improvement loop driven by analyst feedback are covered hands-on. The output is an anomaly detection prototype adapted to the organisation's data, with monitoring recommendations.
Who Should Attend?
- Data scientists and ML engineers
- IT operations, SRE and observability teams
- Quality, manufacturing and process analytics specialists
- Prerequisite: Python and basic machine learning
Key Takeaways
- Distinguish point, contextual and collective anomalies
- Select and apply an anomaly algorithm suited to the data type
- Evaluate with metrics such as precision@k when labels are scarce
- Tune alert thresholds according to business impact
- Design a process that feeds analyst feedback back into the model
Curriculum
Anomaly Types and Statistical Methods
- Point, contextual and collective anomalies
- Z-score, MAD and IQR-based detection
Unsupervised ML Algorithms
- Isolation Forest, LOF, One-Class SVM
- Comparing algorithms with PyOD
Autoencoders and Deep Approaches
- Anomaly scores from reconstruction error
- High-dimensional and mixed data types
Anomalies in Time Series
- Residual analysis after removing seasonality
- Change-point detection and streaming data
Evaluation and Alert Management
- Precision@k, threshold tuning and alert fatigue
- Analyst feedback loop and monitoring
Other Courses in This Category
-
Applied Machine Learning
End-to-end machine learning with Scikit-learn, XGBoost, LightGBM, and CatBoost, covering feature engineering, Optuna hyperparameter tuning, imbalanced data, and SHAP/LIME explainability.
- Live Course
- 3 modules
-
Behavioural features, gradient boosting, anomaly scores and hybrid rule-model setups for payment, insurance and account fraud, with hands-on work on extreme class imbalance, delayed labels and operational threshold management.
- Live Course
- 5 modules
-
Variable analysis with WoE/IV, logistic regression scorecards, gradient boosting alternatives, reject inference and score calibration for credit, insurance and supplier risk, building explainable and auditable risk scores.
- Live Course
- 5 modules
-
Churn, purchase and cross-sell propensity models covering target definition, customer behaviour features, survival analysis, uplift modelling and campaign targeting, turning predictions into measurable retention and sales actions.
- Live Course
- 5 modules
-
Collaborative filtering, content-based and hybrid recommenders, matrix factorisation, implicit feedback, cold-start strategies and ranking metrics for building personalised product and content recommendation engines.
- Live Course
- 5 modules
-
Failure probability and remaining useful life (RUL) prediction from sensor, telemetry and maintenance logs, with hands-on signal feature engineering, health indices, early-warning thresholds and maintenance cost optimisation.
- Live Course
- 5 modules
Get in touch
