Research data for: Modeling behavioral ambiguity with multi-prototype representations for household characteristic identification and demand-side customer segmentation from smart-meter data
Description
This dataset contains author-created derived and aggregate research results supporting the associated article. It includes task definitions and class counts; taskwise and split-level predictive metrics; prespecified paired statistical comparisons; local representation-geometry summaries; aggregate temporal profiles and behavioral-crossover analyses; recurring-day archetype results; prototype-mode, stability and vulnerability summaries; and constrained-budget target-screening results. The files support numerical verification, secondary analysis and replotting of the reported aggregate findings. The restricted Irish CER source smart-meter and survey data, household identifiers, household-level predictions or embeddings, project code, model checkpoints and trained weights are not included. The source data remain available by request from ISSDA under DOI 10.7929/ISSDA/BX59EU. Because this is a derived-results package rather than an end-to-end reproduction package, the manuscript and documentation should be consulted for the full analytical protocol.
Files
Steps to reproduce
1. Read README_FIRST.txt and 00_Documentation/README_RESEARCH_DATA.txt. 2. Use MANUSCRIPT_DATA_CROSSWALK.csv to locate the files supporting a manuscript item. 3. Consult DATA_DICTIONARY.csv and the folder-specific README files for row units, keys and metric definitions. 4. Inspect or analyze the individual CSV files; all principal empirical figures and tables are supported by machine-readable aggregate or split-level data. 5. Verify file integrity with SHA256SUMS.txt when required. 6. End-to-end retraining requires separately authorized access to the CER source dataset and is outside the scope of this deposit because the source data and project code are not redistributed.
Institutions
- Bangladesh University of Engineering and TechnologyDhaka Division, Dhaka