Predictive Model Generation Interface for Unstructured Healthcare Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users without detailed knowledge of statistics face challenges in quickly and easily generating and validating predictive models, particularly when dealing with healthcare data that includes unstructured free-text clinical documentation.
Innovation Solution
A method involving the creation of inclusion and exclusion sets from datasets, using natural language processing to identify unique content elements, and sorting them by absolute percentage difference, allowing for the generation of predictive models without requiring explicit encoding of user-specified criteria.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional predictive model tools are used, then model generation capability is achieved, but user accessibility deteriorates due to requiring sophisticated statistical knowledge
Solution Approach 1:
The patent introduces an intermediary processing layer that automatically performs statistical analysis between the user's simple criteria input and the predictive model generation. This intermediary layer handles the complex statistical computations, frequency calculations, and model training processes, allowing users to interact with the system through intuitive, non-technical interfaces while maintaining high model generation capability
Solution Approach 2:
The system performs self-service by automatically conducting statistical analysis, feature selection, and model generation without requiring user expertise. The system autonomously calculates frequencies, identifies significant features, selects appropriate models, and validates results, enabling users with minimal statistical knowledge to generate reliable predictive models
2Measurement precision
If comprehensive data analysis is performed to identify relevant features, then model accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-calculating and storing frequency statistics for all content elements before model generation. The system pre-processes the dataset to compute occurrence frequencies, creates lookup tables, and prepares statistical summaries in advance, so that during actual model generation, feature identification can quickly reference pre-computed values rather than performing full statistical analysis from scratch
Solution Approach 2:
The system dynamically adjusts analysis parameters such as significance thresholds, minimum frequency cutoffs, and feature selection criteria based on the specific dataset characteristics and user requirements. By changing these parameters adaptively, the system optimizes the balance between comprehensive feature analysis and processing efficiency for different scenarios
Data Source
AI summary
A dataset is searched using inclusion set criteria to produce an inclusion set and exclusion set criteria to produce an exclusion set. A set of unique content elements is identified from the inclusion set and the exclusion set. Metrics are derived from the inclusion set, exclusion set, and set of unique content elements, such as a measure, for each unique content element, of the absolute value of the difference between the percentage of records in the inclusion set containing the unique content element and the percentage of records in the exclusion set containing the unique content element. The unique content element set may be sorted and displayed in decreasing order of the above-referenced absolute value. The content element set may be filtered. Individual content elements may be excluded from the content set. A predictive model may be generated based on the resulting version of the content element set.