Перейти к основному содержимому

Qlik Predict - Create the Predictive Model

Эскиз видео
In this part of the Do More with Qlik – Tips and Tricks Edition series, Mike Tarallo walks you through the complete process of creating a machine learning experiment using Qlik Predict — Qlik's no-code AutoML tool. You'll learn how to upload data, configure predictions, deploy models, and generate predictions based on real-time business data. This foundational tutorial also touches on best practices and introduces key concepts like SHAP explainability, training vs. apply datasets, binary classification, and deployment configuration — all within the Qlik Cloud environment. Perfect for Qlik users looking to harness ML without writing a single line of code. 📢 00:00 – Intro & Overview Welcome and outline of the Qlik Predict video series. 🧠 00:13 – What is Qlik Predict? Intro to AutoML in Qlik and what it offers. ⚙️ 01:03 – Qlik Sense & Role Requirements User permissions needed to run ML experiments. 🆕 01:25 – Creating a New Experiment Setting up a fresh ML experiment in Qlik Cloud. 📁 02:51 – Uploading Training Data Loading historical data (CSV) for model training. 🎯 03:21 – Target Selection (Binary Classification) Choosing the target feature: “Ship Late”. 🚀 04:44 – Running the Experiment Running model training and optimization automatically. 🏆 05:52 – Reviewing Model Scores Looking at F1 scores and selecting the best model. 🧪 06:41 – Running Multiple Experiments (Optional) Testing different parameters or retraining models. 📦 07:23 – Deploying the Model Publishing the chosen model for prediction use. 🔁 08:00 – Apply Dataset Setup Uploading current business data to use with the model. 🔗 08:42 – Schema Matching & Data Sync Ensuring apply dataset matches training schema. 🔮 09:40 – Creating the Prediction Dataset Generating yes/no predictions based on new data. 📊 10:08 – Including SHAP Explainability Data Choosing to include explainable ML insights. 🗂 10:57 – Reviewing Generated Files Exploring deployments, predictions, and data outputs. 📱 11:48 – Preparing for Predictive App Overview of files created and preview of next steps.