N-BaIoT processed for anomaly detection

The original data comes from the work of Meidan et al. [1]. It was preprocessed in this setting for comparative analysis of anomaly detection. The following steps have been taken as preprocessing: (1) five devices have been selected: Danmini doorbell, Ecobee thermostat, Philips baby monitor, Provision security camera, Samsung webcam, (2) for each botnet, the malicious traffic of all five behaviour types have been merged, (3) for each device and botnet combination, malicious requests have been sampled to comprise 5% of the final dataset.

[1] Meidan, Y., Bohadana, M., Mathov, Y., Mirsky, Y., Shabtai, A., Breitenbacher, D., & Elovici, Y. (2018). N-baiot—network-based detection of iot botnet attacks using deep autoencoders. IEEE Pervasive Computing, 17(3), 12-22.

本数据集的原始数据源自Meidan等人[1]的研究成果。本研究在此研究框架下对其进行预处理，以开展异常检测的对比分析。已完成以下预处理步骤：
(1) 选取5款设备：丹米（Danmini）门铃、Ecobee恒温器、飞利浦（Philips）婴儿监护仪、Provision安防摄像头、三星（Samsung）网络摄像头；
(2) 针对每类僵尸网络（botnet），合并其全部5种行为类型的恶意流量；
(3) 针对每款设备与僵尸网络的组合，对恶意请求进行采样，使其占最终数据集总规模的5%。

[1] Meidan, Y., Bohadana, M., Mathov, Y., Mirsky, Y., Shabtai, A., Breitenbacher, D., & Elovici, Y. (2018). N-baiot——基于网络的物联网（Internet of Things, IoT）僵尸网络攻击检测方法：采用深度自编码器的方案. 《IEEE普适计算》(IEEE Pervasive Computing), 17(3), 12-22.