IP2FL: Interpretation-Based Privacy-Preserving Federated Learning for Industrial Cyber-Physical Systems
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
The expansion of Industrial Cyber-Physical Systems (ICPS) has introduced new challenges in security and privacy, highlighting a research gap in effective anomaly detection while preserving data confidentiality. In the ICPS landscape, where vast amounts of sensitive industrial data are exchanged, ensuring privacy is not just a regulatory compliance issue but a critical shield against industrial espionage and cyber threats. Existing solutions often compromise data privacy for enhanced security, leaving a significant void in protecting sensitive information within ICPS networks. Addressing this, our research presents the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">IP2FL</i> model, an Interpretation-based Privacy-Preserving Federated Learning approach tailored for ICPS. This model combines Additive Homomorphic Encryption (AHE) for privacy with advanced feature selection methods and Shapley Values (SV) for enhanced explainability. The proposed solution mitigates privacy concerns in federated learning, where traditional methods fall short due to computational constraints and lack of interpretability. By integrating AHE, the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">IP2FL</i> model minimizes computational overhead and ensures data privacy. Our dual feature selection approach optimizes system performance while incorporating SV to provide critical insights into model decisions, advancing the field towards more transparent and understandable AI systems in ICPS. The validation of our model using ICPS-specific datasets demonstrates its effectiveness and potential for practical applications.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.012 | 0.001 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.000 | 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 it