<aside> 📌 PATH: 360° Customer View → Predictive Models

</aside>

How to train new ML Model:

Prepare the dataset according to requirements

To train an ML model in Bind Genius, please prepare historical datasets following the guidelines provided in:

How to Prepare Training Dataset with Metrics

How to Prepare Training Dataset with Targets

Add training datasets with metrics and targets

Before creating a new model, please add historical datasets with relevant metrics and targets by following our guide:

How to Add Dataset

Press “Add new model”

Press the button “Add new model” above the table with models.

Fill in the required fields

1. Model Name - Create a descriptive name to your model for easy identification..
2. Learning Cycles - Specify how many training attempts the model should make. The best iteration will be automatically selected for testing and future use.

<aside> 📌

Choose 1 learning cycle for the quickest results, suitable for rapid testing and experimentation.

For more in-depth analysis and improved outcomes, we recommend trying multiple learning cycles.

</aside>

1. Model Type - Choose which type of model will help you answer your business question.

<aside> 📌

• Binary models - are used for questions with two possible outcomes (e.g., 0/1, Yes/No, True/False). This model type is ideal for answering questions like: "Will event X happen?" and determining the probability of that outcome.
• Multiclassification - is used for questions where you need to select one outcome from more than two possible options. Example: "Which offer from the list will the customer respond to, and with what probability?”
• Regression models- are used to predict continuous values. Example: "What will the customer's Customer Lifetime Value (CLV) be?” </aside>