Analyzing Federated Learning Aggregation and Distributed Personalization Algorithms Towards Understanding Users’ Residential Electric Load Patterns
Bibliographic record
Abstract
Privacy concerns arise from sharing electric load data. For instance, this data can be hijacked and deductions can be made about building occupancy. Anonymized data does not fully solve the issue, as there have been several successful attempts of re-identifying individuals from anonymized data. Federated Learning (FL) involves training a global model among several clients at the edge without sharing data, but instead, sharing model weights. Besides training a global model to reduce the universal error, it is important to focus on reducing the local error for each client. Personalization helps improve local model convergence without affecting the global model learning process. Therefore, individuals can have a more accurate understanding of their energy usage behavior. It also reduces the number of FL rounds needed for training, which reduces load on the smart grid edge computing resources, communication network, and therefore reduces costs. Previously, research has only looked at stochastic gradient descent (SGD) and Adam for fine-tuning clients’ local models at the edge in the process of FL. This research simulates different FL environments to explore electric load forecasting using different FL aggregation algorithms at the server, as well as seven local optimizers for FL personalization (SGD, Adam, Adagrad, Adadelta, Adamax, Nadam, and RM-SProp). An analysis of the different local optimizer algorithms is also presented.
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 | 0.004 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".