Course Objective & Scope
For time-dependent data such as sales, traffic, energy use or cash flow, a random train/test split gives misleading results. The programme covers trend and seasonality decomposition, stationarity, statistical models such as ARIMA/SARIMA and ETS, ML approaches such as LightGBM with lag features, and libraries including statsforecast, sktime and Darts.
Rolling-origin backtesting, metrics such as MASE and sMAPE, and prediction intervals via conformal methods are practised hands-on. Participants produce a model comparison report on data resembling their organisation's series.
Who Should Attend?
- Data scientists and forecasting analysts
- Finance, planning and operations analytics teams
- Energy, telecom and retail data teams
- Prerequisite: Python and basic statistics
Key Takeaways
- Decompose a series into trend, seasonality and noise
- Build and compare statistical and ML-based forecasting models
- Measure performance honestly with rolling-origin backtesting
- Include exogenous drivers such as holidays and campaigns
- Report uncertainty with prediction intervals
Curriculum
Time Series Analysis
- Trend, seasonality, STL decomposition
- Stationarity, ACF/PACF and autocorrelation
Statistical Models
- ARIMA, SARIMA, ETS and Theta
- Modelling many series with statsforecast
Forecasting with Machine Learning
- LightGBM with lag, rolling and calendar features
- Recursive and direct multi-step strategies
Exogenous Variables and Hierarchical Series
- Modelling holiday, price and campaign effects
- Hierarchical reconciliation
Backtesting and Uncertainty
- Rolling-origin evaluation, MASE, sMAPE
- Conformal prediction intervals and a model comparison report
Other Courses in This Category
-
Detecting unusual behaviour in unlabelled data with statistical methods, Isolation Forest, LOF, autoencoders and time series anomaly techniques, including threshold setting, alert fatigue and evaluation strategies.
- Live Course
- 5 modules
-
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
Get in touch
