Online Lead Scoring Training Activity
Averroes online lead scoring training activity on UiPath
Averroes online lead scoring training activity enables you to train a lead scoring model based on your data directly from UiPath without any required coding.
In this tutorial, we will go with detailed steps on how to use this activity:
- Assuming you already have a project, create a new sequence:
- If you haven’t downloaded the packages, download them. Afterwards, search for online lead scoring training activity.
First, let's explore the details of this activity:
Authentication: On the right-hand side (properties panel); you should fill in your Averroes account info: Email and Password without quotations.
Input:
- Dataset File: path of the file contains the data you want to train the online lead scoring model on it; data should be in Tabular (structured in CSV format).
- Lead Status Column: column name that presents the lead status in your dataset file; should encapsulate in double quotations.
- Training Time: in minutes; as an integer.
Output:
- Message: a message that indicates the status of the training; success, failure, etc.
- Performance Results: a message contains the performance metrics of the generated model (accuracy, f1-score, precision, and recall).
- Fill the email and the password in the authentication section
- To prevent errors that may be produced from using an absolute path, it's better to use the built-in UiPath activity "Select File" which is more user friendly.
- As a UiPath user, it's assumed that creating new variables to store output values is familiar to you. Otherwise, follow these steps:
- Add the "Select File" activity, and select it.
- From its properties window, click on the "+" icon beside selectedFile in Output section, then choose "Create Variable". Write a unique variable name (ex; trainingFilePath) to store the selected file path, in order to use as input to the Dataset File in online lead scoring training activity.
- Fill in the other inputs.
- Store the online lead scoring activity output messages in variables, in the same way we did previously.
- Use the built-in UiPath activity "Message Box" to show each message of the online lead scoring training activity.
- Once you hit Run, the sequence will start execution by showing a pop-up window asking you to select the dataset file:
- If you're in Debug mode, log messages will appear on the right-hand side panel during the execution.
- After the execution is completed successfully, the following two messages will be shown:
- That's great, now you are aware of your generated model performance, but check your email (which was used to create your Averroes account) to find more details.
- Sample email:
- Your model is identified by "Config ID" and "Model ID" values sent in the email. These need to be provided in the prediction activity to make predictions. More details are available in the next section.
- From the dataset link sent in the email, you can reach the training dataset used for this model, which provides many features:
- Understand and get more insight into your data via Preview and Visualization tabs.
- Perform different training executions based on different settings.
- Compare different results of the executions and choose the best to use in prediction.
- From the Prediction tab, for the selected model, you can test the prediction of a single sample.
- Generated Model-ID and Config-ID from the integration section in the Prediction tab; so you can copy them from there directly to use in UiPath Online Lead Scoring prediction activity.
Following is a set of screenshots of these amazing features:
Preview tab
Visualization tab
Train tab
Train tab - test model
Prediction tab
Prediction tab - UiPath integration section