Pith. sign in

REVIEW 6 cited by

KAN-AD: Time Series Anomaly Detection with Kolmogorov-Arnold Networks

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2411.00278 v4 pith:VTFNRQEB submitted 2024-11-01 cs.LG

Quan Zhou , Changhua Pei , Fei Sun , Jing Han , Zhengwei Gao , Dan Pei , Haiming Zhang , Gaogang Xie
show 1 more author
Jianhui Li
This is my paper
classification cs.LG
keywords serieslocaltimetsaddetectiondisturbanceskan-adanomaly
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Time series anomaly detection (TSAD) underpins real-time monitoring in cloud services and web systems, allowing rapid identification of anomalies to prevent costly failures. Most TSAD methods driven by forecasting models tend to overfit by emphasizing minor fluctuations. Our analysis reveals that effective TSAD should focus on modeling "normal" behavior through smooth local patterns. To achieve this, we reformulate time series modeling as approximating the series with smooth univariate functions. The local smoothness of each univariate function ensures that the fitted time series remains resilient against local disturbances. However, a direct KAN implementation proves susceptible to these disturbances due to the inherently localized characteristics of B-spline functions. We thus propose KAN-AD, replacing B-splines with truncated Fourier expansions and introducing a novel lightweight learning mechanism that emphasizes global patterns while staying robust to local disturbances. On four popular TSAD benchmarks, KAN-AD achieves an average 15% improvement in detection accuracy (with peaks exceeding 27%) over state-of-the-art baselines. Remarkably, it requires fewer than 1,000 trainable parameters, resulting in a 50% faster inference speed compared to the original KAN, demonstrating the approach's efficiency and practical viability.

Discussion (0). Sign in to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Trio: Learning Time-Series Forecasting with Temporal-Spatial-Sample Attention and Structural Causal Priors

    cs.LG 2026-06 unverdicted novelty 6.0 of 10

    Trio proposes Temporal-Spatial-Sample attention and a TS-SCM synthetic data generator to improve multivariate time-series forecasting by reusing historical patterns and structural priors.

  2. DMF2Mel: A Dynamic Multiscale Fusion Network for EEG-Driven Mel Spectrogram Reconstruction

    cs.SD 2025-07 reject novelty 5.0 of 10

    DMF2Mel, a dynamic multiscale fusion network, reports the best mel spectrogram reconstruction scores on SparrKULee, though test-set hyperparameter tuning makes the comparison unreliable.

  3. Fourier-KAN-Mamba: A Novel State-Space Equation Approach for Time-Series Anomaly Detection

    cs.LG 2025-11 unverdicted novelty 4.0 of 10

    Fourier-KAN-Mamba combines Fourier features, KAN nonlinearities, and Mamba state-space modeling with a gating mechanism and reports better anomaly detection performance than prior methods on the MSL, SMAP, and SWaT be...

  4. Kolmogorov-Arnold Networks-based GRU and LSTM for Loan Default Early Prediction

    cs.LG 2025-07 conditional novelty 4.0 of 10

    A GRU-KAN hybrid predicts loan defaults 3 and 8 months ahead with reported accuracies of 92% and 88% on a balanced Freddie Mac test set, outperforming LSTM, GRU, and attention baselines.

  5. TPA-AD: A Two-Stage Pseudo Anomaly-Guided Method for Bearing Time-Series Anomaly Detection

    cs.LG 2026-06 unverdicted novelty 3.0 of 10

    TPA-AD generates boundary-near pseudo-anomalies via reconstruction, applies contrastive learning, and uses KNN to score anomalies in bearing time series with only normal training samples.

  6. A Practitioner's Guide to Kolmogorov-Arnold Networks

    cs.LG 2025-10 accept novelty 3.0 of 10

    A systematic review of Kolmogorov-Arnold Networks that maps their relation to Kolmogorov superposition theory, MLPs, and kernels, examines basis-function design choices, summarizes performance advances, and supplies a...

Pith tools