You have received a training-serving skew alert from a Vertex AI Model Monitoring job that is active in a production environment. In response, you have retrained the model using more up-to-date training data and subsequently redeployed it to the Vertex AI endpoint. Despite these actions, you continue to receive the same alert. What step should you take to address this situation?