Qlik Predict Tips - Tip #2 - Data must contains outcomes

In this video, Mike Tarallo from Qlik shares Tip #2 in the Qlik Predict series—focusing on the importance of selecting the right data for building effective machine learning models. If you're using Qlik Predict, it’s critical to understand why traditional analytics data may not be suitable for machine learning and how to choose data that leads to accurate predictions. This video will help you identify the key data ingredients and ensure your data sets are properly structured to support machine learning workflows in Qlik. ⏱️ 00:00 – Introduction Mike introduces the focus of the video: the type of data needed for Qlik Predict. ⏱️ 00:10 – Tip #2: Using the Right Data An overview of what kind of data should be used when creating a new ML experiment in Qlik Predict. ⏱️ 00:14 – Selecting Data from Qlik Catalog Explains how data is selected from the Qlik Catalog and clarifies that not all registered data is suitable. ⏱️ 00:23 – Why Typical Analytics Data Isn’t Enough Details why standard analytics/reporting data (like snapshots with dimensions and measures) generally doesn’t work for machine learning. ⏱️ 00:38 – The Need for Outcomes and Features Discusses the importance of having a defined outcome and pre-existing features in your dataset. ⏱️ 01:00 – Learning Patterns for Prediction Explains how models learn patterns from historical data to predict future outcomes. ⏱️ 01:06 – Applying the Model How models are applied to data to generate predictions based on learned patterns. ⏱️ 01:10 – Importance of Properly Prepared Data Emphasizes selecting catalog data that includes all the necessary components for ML training. ⏱️ 01:18 – Sneak Peek: Tip #3 - Architecting the Data Set Teases the next tip in the series, which will focus on how to properly structure the dataset for predictive modeling.