You are tasked with developing a machine learning model that will handle continuous streaming data from multiple vendors. The goal is to use BigQuery ML for creating the model and Vertex AI for hosting it, facilitating near-real-time data processing. Given the possibility of invalid data values within the stream, what steps should you take to achieve this objective? | Google Professional Data Engineer Quiz