An Interpretable Sparse Fourier-Based Convolutional Neural Network (SpFFT-CNN) for Eccentricity Faults Classification
Bibliographic record
Abstract
This paper presents an interpretable frequency-domain approach for fault classification in synchronous machines using a Sparse Fourier-based Convolutional Neural Network (SpFFT-CNN). The method extends a time-domain convolutional neural network by integrating learnable spectral weights that operate directly in the frequency domain. These weights adaptively emphasize the harmonic components that contribute most to class separation, allowing the network to identify discriminative frequency regions for different fault conditions. The model was applied to current signals containing healthy and eccentric faults under three preprocessing configurations: with DC (direct current) component, without DC, and without mean. The harmonic spectra learned in each case reveal consistent frequency patterns that remain stable across all the model variants, confirming that the network captures genuine spectral characteristics. The higher harmonics consistently define mixed eccentricity, the lower harmonics dominate dynamic eccentricity, and static eccentricity maintains an overlap that is more pronounced when DC is retained. The results demonstrate that the SpFFT-CNN not only achieves a perfect fault classification but also provides a transparent interpretation of the underlying frequency structure. This framework bridges signal-based analysis and deep learning by linking classification accuracy directly to harmonic interpretability.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| 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".