Models
Dflux offers pre-trained machine learning models to facilitate predictions on analyzed data, contributing to informed decision-making. This functionality provides users with a resourceful set of models that can be employed for various analytical purposes, enhancing the platform's capability to generate valuable insights from the data under examination. This module is designed to empower users with an additional layer of analytical depth, enabling them to leverage pre-trained models efficiently within their data science workflows.
ML modeling
The ML modeling of Dflux involves four simple steps:
-
Preprocessing
where you will further process the data generated from the query according to their requirements and goals.
-
Select a model
where the user can select from different analysis methods. Each analysis method will provide algorithms that the user can select - either
single or multiple algorithms. The user then has to select a Target Variable that the model should be applied to before clicking on the
<Run Model> button.
-
Execute the model
where the model is executed based on the selected methods, algorithms, and parameters, giving a thorough description of model details along with the accuracy levels and predictions.
-
Model analysis
this step, being the final step in ML modeling, provides a comprehensive analysis of the findings. The user can create visualizations for these ML models by clicking on Create Visualization. This will take them to the Visualization window. These visualizations can be added to the dashboard and can be downloaded or shared with others.
Once the query is saved, you can apply machine learning models by clicking on New Model and proceeding with selecting the query on which you want to apply this model.
Once the preprocessing and selecting a model is done, the model is executed and a model summary is provided along with other details like accuracy, precision, and other model metrics. You can create charts for your ML models and add them to your dashboards easily.