Staged ML Risk Assessment for Lower Computer Resource Use
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing risk assessment methods consume excessive computer resources due to the need to process and transmit entire questionnaires, which is not optimal in terms of processing and remote client-server connections.
Innovation Solution
A computer-implemented method that uses incremental machine learning models to assess risk, where demographic data is initially evaluated, and only if thresholds are not met, additional context-specific data is gathered and evaluated by subsequent models, thereby minimizing unnecessary resource use.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the entire questionnaire is processed and transmitted for risk assessment, then the assessment completeness is improved, but the computer resource consumption increases
Solution Approach 1:
The risk assessment process is divided into multiple stages: initial demographic data processing, intermediate risk score calculation, and conditional detailed assessment. Only subjects who fail the initial threshold assessment proceed to the next stage, avoiding unnecessary processing of complete questionnaires for low-risk subjects.
Solution Approach 2:
The system performs preliminary risk assessment using demographic data before requiring completion of the entire questionnaire. This preliminary action filters out subjects who can be approved or rejected based on initial criteria, preventing unnecessary transmission and processing of additional questionnaire data.
2Measurement precision
If supplemental context-related subject data is collected and processed for all subjects, then the assessment accuracy is improved, but the data transmission and processing load increases
Solution Approach 1:
Different levels of data collection are applied to different subjects based on their initial risk assessment. Subjects who pass the initial threshold receive basic demographic assessment, while only those who fail proceed to supplemental data collection, creating localized data quality appropriate to each subject's risk profile.
Solution Approach 2:
The system collects complete supplemental context-related data only when necessary (for subjects failing initial assessment), rather than for all subjects. This partial action approach ensures data accuracy where needed while avoiding excessive data collection and transmission for subjects who can be assessed with minimal data.
3Measurement precision
If multiple machine learning models are sequentially applied, then the risk assessment precision is improved, but the processing time increases
Solution Approach 1:
The multi-model assessment is segmented into conditional stages. The first machine learning model processes demographic data initially, and only subjects failing this first model proceed to the second model with supplemental data. This segmentation ensures multiple models are applied only when necessary, reducing overall processing time while maintaining precision for complex cases.
Solution Approach 2:
The first machine learning model performs preliminary filtering before the second model is engaged. This preliminary action using demographic data alone quickly identifies subjects who need further assessment, preventing unnecessary execution of the more computationally intensive second model for all subjects.
Data Source
AI summary
A method for managing computer resource use in computer-implemented risk assessment of a subject in respect of a context. Demographic subject data for the subject is passed to a first trained machine learning model trained on first context-specific historical risk outcomes correlated with historical demographic data corresponding to the demographic subject data. If a first threshold assessment from the first trained machine learning model is passed, the subject is approved. Responsive to failing the first threshold assessment, supplemental context-related subject data for the subject, in addition to the demographic subject data, is passed with the demographic subject data to a second trained machine learning model trained on second context-specific historical risk outcomes correlated with the historical demographic data and with historical context-related data corresponding to the supplemental context-related subject data. If a second threshold assessment from the second trained machine learning model is passed, the subject is approved.