Forge AI Configuration Documentation

Overview

Forge AI by FojiSoft is a powerful no-code platform that enables users to configure and deploy machine learning models with ease. This guide will walk you through the steps to create a new configuration in Forge AI, including setting parameters for the task, data selection, and training options.

Steps for Configuring Forge AI

1. Initiate Configuration
   • Click the New Configuration button located at the top right corner of the Forge AI interface.
2. Configuration Details
   • Name: Enter a unique name for your configuration.
   • Task: Select the type of task for your model. Options include:
     • Regression
     • Classification
   • Model Retention: Select the criteria for retaining models.
3. Data Selection
   • Data Set: Choose a data set from the drop-down menu of created data sets. Refer to the data set documentation for more details.
   • Range: Specify the number and select the appropriate unit (seconds, minutes, hours, days, weeks).
   • Steps: Enter the number and select the appropriate unit (seconds, minutes, hours, days, weeks, months, years).
   • Input Fields: Select all the fields that apply for input.
   • Output Field: Select the applicable field for output.
4. Training Configuration
   • Mode: Choose between:
     • On-Demand
     • Automatic
   • Duration: Enter the number and select the appropriate unit (seconds, minutes, hours).
   • Training Interval: (Only if Automatic is selected) Fill in the number and select the unit (second, minute, hour, day, week).
5. Finalize Configuration
   • Once all parameters are set, click the Create button at the top right to save your new configuration.

Detailed Explanation of Parameters

• Name: This helps in identifying the specific configuration among others.
• Task:
  • Regression: For predicting continuous values.
  • Classification: For categorizing data into predefined classes.
• Model Retention: Specifies the criteria and conditions under which models are retained.
• Data Selection:
  • Data Set: Lists available data sets which have been created and are ready for selection.
  • Range & Steps: Define the temporal aspect of data, which is crucial for time-series analysis.
  • Input Fields: These are the features used to train the model.
  • Output Field: This is the target variable the model will predict.
• Training Configuration:
  • Mode:
    • On-Demand: Training is initiated manually.
    • Automatic: Training is scheduled at regular intervals.
  • Duration: The length of each training session.
  • Training Interval: Specifies how often the automatic training should occur.

Example Configuration

Here’s an example of a configuration setup for predicting machine maintenance needs:

1. Initiate Configuration:
   • Click New Configuration.
2. Configuration Details:
   • Name: Predictive Maintenance Model
   • Task: Classification
   • Model Retention: Keep up to 5 models
3. Data Selection:
   • Data Set: Maintenance Data Set
   • Range: 24 (hours)
   • Steps: 1 (day)
   • Input Fields: Temperature, Vibration, Pressure
   • Output Field: Maintenance Required
4. Training Configuration:
   • Mode: Automatic
   • Duration: 2 (hours)
   • Training Interval: 1 (day)
5. Finalize Configuration:
   • Click Create.

By following these steps, you can effectively set up and configure your Forge AI environment to meet your specific needs. For further details on each parameter and more advanced configurations, refer to the additional resources and documentation provided by FojiSoft.