Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Matrix Factorisation for Scalable Energy Breakdown
Authors: Nipun Batra, Hongning Wang, Amarjeet Singh, Kamin Whitehouse
AAAI 2017 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate our approach on 516 homes from a publicly available data set and find it to be more effective than five baseline approaches that either require sensing in each home, or a very rigorous survey across a large number of homes coupled with complex modelling. |
| Researcher Affiliation | Academia | 1IIIT Delhi, India, 2 University of Virginia, USA |
| Pseudocode | No | The paper provides mathematical formulations for its optimization problem (Equations 1 and 2) but does not include any pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | Yes | We used the results from this study to produce an open prototype of the system: a web application2 that can potentially provide energy breakdown for millions of homes across the US. 2https://github.com/nipunbatra/scalable-energy-breakdown-webapp. Our entire implementation, experiments and analysis can be found on Github (URL not mentioned for anonymity). |
| Open Datasets | Yes | We use the publicly available Dataport (Parson et al. 2015) data set for evaluation. Dataport is the largest5 public data set for household energy data. |
| Dataset Splits | Yes | For both Gemello and MF, we use a nested leave-one-out cross-validation strategy. The inner loop is used to fine-tune the parameters. The outer loop is used for prediction of energy across different appliances for a test home, when all but that home are used in the train set. |
| Hardware Specification | No | The paper discusses hardware related to the problem domain (e.g., smart meters) but does not specify any hardware used for running their experiments, such as CPU/GPU models, memory, or cloud instances. |
| Software Dependencies | No | The paper mentions using 'NILMTK' for baselines but does not provide specific version numbers for NILMTK or any other software components critical for reproducing their own implementation or experiments. |
| Experiment Setup | Yes | Our MF based approach has regularisation (λ), static features to include (area, number of occupants and number of rooms) and the number of latent factors as the tunable parameters. We varied λ in factors of 10 from 10^-3 to 10^2. We used all length-0, 1, 2 and 3 combinations of the 3 static features (<None>, <area>, <#occupants>,. . . <area, #occupants, #rooms>). We varied the number of latent factors from 1 to 10. |