Design of classification models for identifying telecommunication devices

随着信息技术(IT)和电信系统的规模和复杂性不断增长， 支持提供商服务所需的电信和网络设备的类型和数量 已经激增。识别网络中的这些设备是一项具有挑战性的任务，因为它们通常缺乏适当的或 描述其操作特征的准确文档。电信设备分类为 这不是一个微不足道的问题，因为它需要很好地理解不同类型的 以及根据它们的特征特征对它们进行分类的能力。 开发了不同的特征选择算法来识别这些设备特有的特征 这是这项工作的第一步，这是由NSERC的Engage赠款支持的。 该项目的第二阶段旨在为精确设备开发不同的分类器算法 分类。 该项目第二阶段的总体目标是：1)设计能够 处理广泛的特征；2)评估不一致和缺失数据对精度的影响 一种设备的分类。

As information technology (IT) and telecommunication systems continue to grow in size and complexity, the types and the number of telecommunication and networking devices required to support provider's services have surged. Identifying these devices in a network is a challenging task, since they usually lack proper or accurate documentation describing their operational characteristics. Telecommunication device classification is a non-trivial problem, since it requires a good understanding of the features specific to different types of devices along with an ability to classify them based on their feature characteristics. The development of different feature selection algorithms to identify the features specific to these devices was the first step in this exercise, which was supported by an NSERC's Engage grant. The second phase of this project aims to develop different classifier algorithms for accurate device classification. The overall objectives of the second phase of this project are: 1) to design classification models capable of dealing with a wide range of features; 2) to evaluate the effect of inconsistent and missing data on the accuracy of a device's classification.