For this guide, we will be using the Titanic dataset from Kaggle. In a previous tutorial, we covered how you can use that dataset to create a simple predictive model. In this guide, we will be creating a pipeline to automatically update our training data and then use it to re-train the model every day.
For the sake of this guide, the dataset hosted on this URL gets updated on regular intervals and we will create a reducer to directly fetch data from a URL.
On the left-hand menu, click on "New Pipeline". Then create a task to automatically update your TFrame's data.
Create a task to automatically retrain your model using the updated training data.
At the bottom of the page, schedule your pipeline to run every day.