A problem with machine learning (ML) techniques for detecting intrusions in the Internet of Things (IoT) is that they are ineffective in the detection of low-frequency intrusions. In addition, as ML models are trained using specific attack categories, they cannot recognize unknown attacks. This article integrates strategies of cost-sensitive learning and multitask learning into a hybrid ML model to address these two challenges. The hybrid model consists of an autoencoder for feature extraction and a support vector machine (SVM) for detecting intrusions. In the cost-sensitive learning phase for the class imbalance problem, the hinge loss layer is enhanced to make a classifier strong against low-distributed intrusions. Moreover, to detect unknown attacks, we formulate the SVM as a multitask problem. Experiments on the UNSW-NB15 and BoT-IoT datasets demonstrate the superiority of our model in terms of recall, precision, and F1-score averagely 92.2%, 96.2%, and 94.3%, respectively, over ot
Wire Arc Additive Manufacturing (WAAM) has revolutionized the manufacturing paradigm for fabricating medium to large scale metallic parts featuring high buy-to-fly ratios such as aerospace components. As a promising technology for the manufacturing industry, it is necessary to develop an automated WAAM system with high efficiency and low labour cost. Generally, to achieve a fully intelligent WAAM system, the first step is to develop an intelligent weld bead modelling system which is able to provide users with appropriate welding parameters in terms of producing components with high accuracy. Knowledge from many disciplines, such as computer science, material engineering, mechanical engineering, and industrial system engineering, is advantageous to develop such an automated system. Thus, an intelligent bead modelling system was developed by integrating a number of industrial sectors in this study. The bead modelling system includes three critical modules, including data generation modul