Data supporting the publication: Global Anomaly Detection using Feedforward Symmetrical Autoencoder Neuronal Network. Comparison with Other Methods in a Case Study using Real Industrial Data.

The research objective was to evaluate artificial intelligence-based methods for global anomaly detection in a drinking water treatment plant, where data was stored in a timeseries format. The Autoencoder neural network-based method proved superior to 5 other methods with which the study compared to, in terms of precision, recall and F1 score metrics.

本研究旨在评估基于人工智能的方法在饮用水处理厂中的全局异常检测应用效果，该厂的相关数据以时间序列格式存储。实验结果表明，基于自编码器（Autoencoder）神经网络的方法在精确率、召回率与F1分数三项评估指标上，均优于本研究所对比的其余五种方法。