Wire-arc additive manufacturing (WAAM) is an arc-based directed energy deposition approach that uses an electrical arc as a source of fusion to melt the wire feedstock and deposit layer by layer. It’s applicable in fabricating large-scale components. At this stage, there are still some issues that need to be researched deeply, such as manufacturing accuracy control, process parameters optimization, path planning, and online monitoring. Machine learning is a new emerging artificial intelligence technology, which is more and more applied in modern industry. In this study, a machine learning based control algorithm was applied in melt pool width control. To monitor the WAAM process, deep learning algorithms were applied in anomalies recognition. At the same time, machine learning methods were employed to predict the deposited surface roughness during the WAAM process.