# ECG Arrhythmia Recognition on Normal, CHF, Ventricular Arrhythmia, and Atrial Fibrillation Source: [ECG Arrhythmia Recognition on Normal, CHF, Ventricular Arrhythmia, and Atrial Fibrillation]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:54:23.177Z Evaluation of ECG Arrhythmia Recognition performance across multiple conditions including Normal beats, CHF, Ventricular Arrhythmia, and Atrial Fibrillation, reporting Overall Accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | ECG Arrhythmia Recognition | | Dataset (including any stated split) | Normal, CHF, Ventricular Arrhythmia, and Atrial Fibrillation | | Metric | Overall Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 1 of 1 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Overall Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Overall Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Ubeyli, 2009 | 98.06 | Not recorded | 68fa3a011e209131ce8c726a | ### Ubeyli, 2009 Overall Accuracy: 98.06. Result ID: 698b97837a967eddb76fdb55. [Reporting paper]() Recorded evaluation configuration: {"Feature Model":"Eigenvector Method","Classifier":"Recurrent Neural Network"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.