Before moving forward with DataRobot and machine learning, it's important you have a clear understanding of what you want to accomplish. Some important questions to ask yourself: How do we frame our predictive question? Who are our stakeholders? How will the information be used? The community has a ton of resources to help you, including the easy-to-use practitioner worksheets and articles explaining how to set yourself up for success. This index, specifically curated for business analysts, provides quick links to resources that will help you make your plans.
For this tour, we're using a dataset containing the details of past shipments of HIV drugs and supplies to countries around the world to help us predict which future shipments are at risk of arriving after their estimated delivery date. It's critical that we avoid delays which can have a serious effect on human health.
Each row in the dataset is a shipment. Let’s say the non-profit you work for is responsible for these shipments. You’ve been asked to find ways to reduce the number of late shipments and present some options to your leadership. Perhaps you might consider changing suppliers or adjusting transportation modes or routes.
This use case presents a good example for how a company can leverage AI models to improve their decision making. A summary of this use case is provided at DataRobot Pathfinder.
Let's get started!
When you are using the AI Platform Trial (the URL shows app2.datarobot.com ), navigate to this use case ("Predict Whether a Parts Shortage Will Occur") and hit Create project; the data will be uploaded into DataRobot automatically.