Tutorial - Visual Market Basket Analysis

  • Visual market basket analysis

    Suppose a customer of your online store looking at a product from your portfolio. A proven fact is that recommendation of similar products can increase customer satisfaction and consequently also sales of your online store. The problem comes, when many products exist in a single category. In this tutorial is shown how to train a visual receommendation system, which will utilize a visual information for recomendations. This tutorial shows, how to use the Image Analysis Enterprise software for visual recomendation system.

    Please note that this is a tutorial, i.e. a simplyfied version of a previous real BurgSys's project. The objective is not to achieve the best possible accuracy but to provide easy to understand and easy to configure example. Accuracy of the process can be improved significantly by model optimization and more information you can also get at BurgSys training.

Training Data and Process Description

  • Example of training data used for advanced similarity measuring.

  • Prcess, which trains a model. The first part extracts numerical features from images, the Decision Tree Learner trains a model.

  • After some model optimization.

  • The trained model can be deployed to a server and process thousands of images per hour, 24 hours a day.