We hear from many clients that one of the hardest parts of machine learning is closing the feedback loop. This means that models need to be monitored and updated frequently to incorporate the latest data. Watson Machine Learning and Watson Studio allow data scientists and analysts to quickly build and prototype models, to monitor deployments, and to learn over time as more data become available. Performance Monitoring and Continuous Learning enable machine learning models to retrain on new data supplied by the user or another data source. Then, all of your applications and analysis tools which depend on the model are automatically updated as Watson Studio handles selecting and deploying the best model.
In this video, we’ll solve a problem for the City of Chicago using the Model Builder to model building violations. We’ll predict which buildings are most likely to fail buildings inspections. Then, we can intelligently rank the buildings by their likelihood to fail an inspection, saving time and resources for the City and for our inspectors. We’ll start building a model on publicly available data from 2017, starting in September before we introduce October, November, and December data to simulate learning over time.
You’ll need Watson Studio, Watson Machine Learning, and IBM Db2 Warehouse on Cloud services on IBM Cloud to complete this guide on your own.
Free and Lite versions are available. Get the data.