Performance Evaluation of Techniques for Identifying Abnormal Energy Consumption in Buildings
Bibliographic record
Abstract
Energy consumption in buildings has steadily increased. Buildings consume more energy than necessary due to suboptimal design and operation. Apart from retro-fitting, not much can be done with the design of the existing building, but the operation of the building can be improved. Ignoring or failing to fix the faults can lead to problems like the higher cost in excess energy usage or premature component failure. At the same time understanding, identifying, and addressing abnormal energy consumption in buildings can lead to energy savings and detection of faulty appliances. This paper investigates two key challenges found in energy anomaly detection research: 1) the lack of labeled ground truth and 2) the lack of consistent performance accuracy metrics. In the first challenge, labeled ground truth is imperative for training and benchmarking algorithms to detect anomalies. In the second challenge, consistent performance accuracy metrics are crucial to quantifying how well algorithms perform against each other. There exists no publicly available energy consumption dataset with labeled anomaly events. Therefore, we propose two approaches that help in the automatic annotation of the ground truth data from publicly available datasets: a statistical approach for short-term data and a piecewise linear regression method for long-term data. We demonstrate these approaches using two publicly available datasets called Dataport (Pecan Street) and HUE. Using different existing accuracy metrics, we run a series of experiments on anomaly detection algorithms and discuss what metrics can be best used for consistent accuracy testing amongst researchers. In addition, while providing the source code, we also release an anomaly annotated dataset produced by this source code.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.022 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.004 | 0.004 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".