In Overview of Einstein Discovery Part 1 – Mortgage Banking Technology | Origination Strategy Consulting (takefiveconsulting.com) and OVERVIEW OF EINSTEIN DISCOVERY PART 2 – Mortgage Banking Technology | Origination Strategy Consulting(takefiveconsulting.com), we learned the basics of Einstein Discovery. In this article we will dive into the details of how to Create model in Einstein Discovery.
Considerations for Model Creation
- Einstein Discovery requires a minimum of 400 rows containing outcome values to develop a model effectively.
- During model training, Einstein excludes rows lacking outcome values from consideration.
- Leveraging the capabilities of AI and machine learning, Einstein Discovery enables analysis of datasets containing up to 20 million rows of data.
What Is a Model?
A model represents a refined customized equation, built upon a thorough statistical comprehension of historical results, aimed at forecasting future outcomes. In Einstein Discovery, a model encompasses a spectrum of performance metrics, configurations, predictions, and data-driven insights. Einstein Discovery guides users through the process of constructing a model, aligning with the desired outcome for enhancement (the model’s objective), utilizing the accumulated data for analysis (within the CRM Analytics dataset), and adjusting various settings to guide the analysis process and articulate its findings.
How to create a model in Einstein Discovery?
- On the Analytics Studio home page, click Create and select ‘Model’.
On the New Model screen, click Create from Dataset and click Continue
Select the desired dataset you want to analyse, then click Next. Here we are choosing Lead data.
In the “Create Model” screen, specify your goal. The goal determines the desired outcome for analysis and guides the model training process to predict whether you aim to maximize or minimize the outcome.
In this example, your goal is to maximize lead conversion. For I Want to Predict, select ‘Converted’ and, next to Maximize, select ‘1’ since ‘1’ implies converted and ‘0’ implies not converted. Accept all other the default settings and click Next.
In the Configure Model Columns screen, accept the default (Automated), and click Create Model.
Einstein begins the process of data analysis by employing statistical techniques, machine learning algorithms, and artificial intelligence, ultimately crafting a predictive model.
Upon completion, Einstein presents an overview of the model’s performance.
The Model is not yet ready to be deployed. There are issues with the model quality and Einstein has thrown a bunch of Data alerts to be addressed. We will address the alerts in an upcoming article.
To learn more, click Introduction to Einstein Discovery (salesforce.com).
Take Five Consulting is a technology company, based in Virginia U.S., that specializes in the Mortgage Banking vertical especially LOS implementation and application development. Take Five Consulting creates and implement mortgage technology and software specifically for Mortgage Industry.
