Skip to main content

Machine Learning Methods for Time Series

Machine learning methods for time series, worked on energy and supply chain data, in two books. Each chapter pairs a narrative page with a companion notebook, so that every figure and table in the text is computed by code that the reader can run.

Read Volume 1 online Read Volume 2 online Both books

Volume 1. Finding Structure in Time Series

Cover of Finding Structure in Time Series: Clustering, Ranking, Recommendation and Anomaly Detection

Clustering, Ranking, Recommendation and Anomaly Detection. The first volume shows how to find structure in many time series at once: which series behave alike, which ones deserve attention first, which cases are similar to a new one, and which observations do not belong. The worked examples follow the smart-meter signal to decisions on the low-voltage distribution network.

Read Volume 1 online

Part 5. Clustering

  1. Clustering Load Profiles
  2. Validating Clusters
  3. Clustering Metered Households
  4. Network Risk
  5. Customers and Connections

Part 6. Retrieval, Ranking and Recommendation

  1. Foundations of Retrieval, Ranking and Recommendation
  2. Retrieval for Connection Requests
  3. Ranking Transformers and Violations
  4. Recommending Mitigation Measures

Part 7. Anomaly Detection

  1. Foundations of Anomaly Detection
  2. Meter Data Quality and Customer Behaviour
  3. Non-Technical Loss and Changes over Time

Volume 2. Forecasting for Operations

Cover of Forecasting for Operations: From Signals to Decisions

From Signals to Decisions. The second volume covers what the signals allow before a model is chosen, the families of forecasters, and how a forecast is evaluated without fooling oneself.

Read Volume 2 online

Part 8. Forecasting

  1. Signals
  2. Forecasters and Their Evaluation

The web editions grow chapter by chapter, as each is written, run and reviewed.

Built on open-source libraries

The methods are implemented in libraries that can be installed and reused outside the books.

Library Role in the books
Milia Clustering of load profiles, retrieval of similar cases and anomaly detection
Popo Low-voltage power flow and state estimation for network studies
Twiga Point and probabilistic forecasting

See the Software page for installation and documentation.

Status

Both volumes are in progress. The other titles are listed on the Books page.