AI Prediction Model Change Policy Evaluation System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AI prediction systems for object events, such as crime risk in city planning or disease incidence in healthcare, lack the ability to provide actionable insights beyond historical marketing policies, failing to utilize AI prediction effectively for improving Key Performance Indicators (KPIs).
Innovation Solution
A computer system that utilizes a prediction model to generate change policy data by altering feature values based on a feature profiling database, calculates evaluation values for these policies, and presents them as actionable information for achieving specific objectives related to the object, leveraging AI prediction results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If AI prediction is utilized to generate change policy data by altering feature values, then actionable information for improving KPIs is provided, but system complexity increases
Solution Approach 1:
The system segments the complex task of generating actionable insights into distinct functional modules: a prediction model processing unit that handles AI predictions, a change policy generation unit that creates alternative scenarios, and an evaluation unit that assesses effectiveness. This modular segmentation reduces overall system complexity while preserving the ability to generate actionable information.
Solution Approach 2:
The patent introduces a feature profiling database as an intermediary component that stores pre-defined change rules and relationships between feature values. This intermediary structure simplifies the generation of change policy data by providing a standardized framework for altering feature values, reducing the complexity of directly manipulating raw prediction data.
2Loss of information
If change policy data is generated by changing multiple feature values based on feature profiling database, then comprehensive evaluation of actionable strategies is achieved, but calculation time increases
Solution Approach 1:
The system performs preliminary actions by pre-defining change rules and relationships between feature values in the feature profiling database before actual prediction analysis. This pre-processing establishes a framework of valid feature transformations, allowing the system to quickly generate change policy data without performing complex calculations during the actual evaluation phase.
Solution Approach 2:
The patent systematically varies parameter values (feature values) according to pre-defined change rules in the feature profiling database. By changing parameters in structured ways rather than exhaustive searches, the system achieves comprehensive evaluation of actionable strategies while controlling calculation time through efficient parameter sampling.
3Measurement precision
If AI prediction model is used to evaluate change policy data, then accuracy of KPI improvement prediction is improved, but computational resources increase
Solution Approach 1:
The system applies partial action by evaluating only the most promising change policy data generated from the feature profiling database, rather than exhaustively analyzing all possible feature value combinations. The prediction model focuses computational resources on assessing a selective subset of high-potential strategies, maintaining prediction accuracy while reducing overall computational resource consumption.
Data Source
AI summary
Provided is a computer system to present information useful for achieving purposes related to an object by utilizing AI prediction. The computer system manages a prediction model for predicting an object event based on evaluation data and feature profiling database that defines a change rule of each of the plurality of feature values included in the evaluation data, generates change policy data by changing the plurality of feature values included in the evaluation data based on the feature profiling database, calculates an evaluation value indicating effectiveness of the change policy data, and generates display data for presenting the change policy data and the evaluation value as information useful for achieving purposes related to the object.