Analytics Server for Multi-Tenant Predictive Model Porting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions face challenges in identifying relevant predictive variables from large datasets and porting predictive models across tenants in multi-tenant environments, as they rely heavily on manual expert intervention and are computationally expensive, and struggle with handling changes in datasets over time.
Innovation Solution
An analytics server that automatically obtains and analyzes input data to determine predictive analysis parameters, compares and adapts predictive models across tenants by identifying missing, new, or changed parameters, and refines models to ensure data distribution changes are accounted for, enabling efficient model porting and scalable predictive analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual expert intervention is used to select predictive variables and port models, then model accuracy can be maintained, but the process becomes labor-intensive and slow
Solution Approach 1:
The system performs self-service by automatically comparing data schemas between source and target tenants, identifying missing or changed parameters, and adapting predictive models without requiring manual expert intervention. The analytics server executes automated routines that detect schema differences and adjust models accordingly, eliminating the need for data scientists to manually evaluate each tenant's data.
Solution Approach 2:
The patent replaces the mechanical process of manual model porting with an automated computational system. The analytics server uses algorithmic comparisons of data schemas and automated model adaptation techniques to substitute the manual expert workflow, significantly reducing time while maintaining model accuracy through systematic parameter matching and adjustment.
2Productivity
If automated model porting is implemented across tenants, then scalability is improved, but handling schema differences and data distribution changes becomes complex
Solution Approach 1:
The system segments the model porting process into distinct automated stages: schema comparison, parameter identification, model adaptation, and validation. By breaking down the complex task of handling schema differences into manageable segments, the system can systematically address each aspect (missing parameters, new parameters, changed parameters) independently while maintaining overall scalability.
Solution Approach 2:
The patent handles schema complexity by dynamically adjusting parameters during model porting. The system automatically detects parameter differences between tenants and applies parameter changes to adapt the model, including adding missing parameters, removing unnecessary ones, and adjusting parameter values based on target tenant data distributions, thereby simplifying the adaptation process.
3Adaptability or versatility
If comprehensive parameter comparison is performed across tenants, then model adaptability is improved, but computational cost increases
Solution Approach 1:
The system extracts only the essential parameters needed for model porting by comparing data schemas and identifying specifically missing, new, or changed parameters. Rather than performing exhaustive comparisons of all data elements, the analytics server extracts and processes only the relevant parameter differences, reducing computational overhead while maintaining model adaptability.
Solution Approach 2:
The patent applies partial action by performing parameter comparison only where necessary for model porting. The system identifies and processes specifically the parameters that differ between source and target tenants, rather than comprehensively analyzing all parameters. This selective approach achieves sufficient model adaptability with reduced computational cost by focusing on critical parameter differences.
Data Source
AI summary
An analytics server for scalable predictive analysis for analytics as a software service in multi-tenant environment is provided. The analytics server automatically validates portability of a predictive model from a first tenant to a second tenant by comparing value distribution of parameters between data inputs of the first tenant and the second tenant. The analytics server further automatically detects source data changes over a configurable time horizon as relevant to predictive model inputs, by comparing value distribution of parameters between two data inputs from a same tenant separated by a selected time horizon.