ML Model Builder Interface With Artifact Dependency View
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data visualization applications require users to build custom machine learning models outside of the data visualization environment, leading to a disjointed user experience, slow delivery of results, and errors due to misunderstood requirements, which segregates predictive model builders from decision-making teams.
Innovation Solution
Integrate visual analytics with predictive analytics by allowing users to create and deploy no-code machine learning models directly within the data visualization interface, enabling collaboration among team members with and without data science backgrounds to train, understand, and operationalize models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If users build custom ML models outside the data visualization application, then model customization and flexibility are improved, but user experience coherence and collaboration efficiency deteriorate
Solution Approach 1:
The patent merges the data visualization application with ML model building capabilities by integrating a model builder interface directly into the visualization environment. This allows users to construct, train, and deploy ML models within the same application where they visualize data, eliminating the need to switch between separate tools and maintaining contextual continuity throughout the analytics workflow.
Solution Approach 2:
The data visualization application is transformed into a multi-functional platform that simultaneously performs data visualization, data preparation, ML model building, model training, and model deployment. This universal platform approach allows a single application to handle the entire analytics pipeline from data exploration to predictive modeling, improving both user experience coherence and operational efficiency.
2Adaptability or versatility
If users build custom ML models outside the data visualization application, then model flexibility is improved, but delivery speed and collaboration efficiency deteriorate
Solution Approach 1:
By combining model building and deployment capabilities within the data visualization application, the patent eliminates time-consuming data transfer and model export processes. Users can immediately deploy models after training without leaving the application environment, significantly accelerating the delivery pipeline from data insights to deployed predictions.
Solution Approach 2:
The application performs preliminary data preparation and validation steps automatically within the model building workflow, preparing data in advance for model training. This preliminary processing reduces the time required for model deployment and ensures data quality requirements are met before models are trained, improving overall delivery speed.
3Manufacturing precision
If centralized data science teams build all ML models, then model quality and expertise are improved, but scalability and responsiveness to business needs deteriorate
Solution Approach 1:
The patent implements a self-service ML model building capability that empowers business users to independently create, train, and deploy their own predictive models without requiring centralized data science team involvement. The application provides automated model selection, hyperparameter tuning, and performance evaluation, enabling non-experts to build quality models while scaling analytics capabilities across the organization.
Solution Approach 2:
The data visualization application serves as an intermediary platform that bridges business users and ML capabilities. It translates business requirements into appropriate model configurations and automatically handles complex ML operations, allowing users to leverage expert-level model building capabilities without needing deep ML expertise, thus scaling quality model deployment across the organization.
Data Source
AI summary
A computer system receives a user input specifying a first data field of a data source and a modeling objective for the first data field. The computer system automatically executing a machine learning (ML) model training workflow to train a model to predict a first outcome for the first data field based on the modeling objective. While automatically executing the ML model training workflow to train the model, the computer system determines a plurality of artifacts across a plurality of steps of the ML model training workflow. The computer system generates and causes display of an artifact dependency view that shows dependency relationships between the plurality of artifacts, including a plurality of visual representations corresponding to the artifacts and a plurality of connectors, a respective connector connecting two visual representations of the plurality of visual representations, corresponding to two artifacts that have a dependency relationship with each other.