Prediction Model Input Filtering for Out-of-Range Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing prediction models face errors when factor values for prediction input are outside the range of past specimens, and there is difficulty in handling factors with time dependency on the specimen range, leading to inaccurate predictions.
Innovation Solution
A system performs factor selection and filtering processes to ensure that only factor values within the specimen range are used for prediction, excluding those outside the range and reducing the influence of time-dependent factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If factor values outside the specimen range are used for prediction, then the prediction model can handle a wider range of inputs, but large errors occur in prediction values due to extrapolation
Solution Approach 1:
The system performs preliminary actions by accumulating factor values over time to expand the specimen range before prediction is needed. This allows the prediction model to be trained on a broader range of data, reducing the need for extrapolation and improving prediction accuracy when new factor values are input.
Solution Approach 2:
The specimen range is made dynamic and time-dependent rather than fixed. The system continuously updates the specimen range by accumulating new factor values over time, allowing the prediction model to adapt to changing conditions while maintaining accuracy within the updated range.
2Measurement precision
If the specimen range is fixed based on past data, then the prediction model is optimized for that specific range, but factors with time dependency cannot be accurately predicted when values change
Solution Approach 1:
The system transforms the static specimen range into a dynamic, time-dependent range that automatically expands as new factor values are accumulated. This allows the prediction model to maintain optimization for the current specimen range while adapting to changing conditions over time, accurately handling factors with time dependency.
Solution Approach 2:
The system incorporates feedback by continuously monitoring whether input factor values fall within the current specimen range. When values outside the range are encountered, the system uses this feedback to expand the specimen range and retrain the prediction model, ensuring ongoing accuracy as conditions evolve.
3Measurement precision
If more factors are selected to improve prediction accuracy, then the model captures more influences on the prediction target, but the complexity of factor selection and processing increases
Solution Approach 1:
The system performs preliminary factor selection and accumulation before prediction is needed. By pre-identifying relevant factors and accumulating their values over time, the system reduces the complexity of real-time factor selection while maintaining comprehensive coverage of influences on the prediction target.
Data Source
AI summary
A system performs factor selection processing, which includes performing a factor selection operation for selecting one or two or more factors from one or a plurality of factors; factor filtering processing which includes: determining, about each of the one or the plurality of factors before the factor selection processing is performed, whether a factor value for prediction input to a prediction model is within a range of a plurality of factor values for specimen used to identify the prediction model; and excluding a factor, a result of the determination of which is false, and outputting a factor not excluded; and prediction processing which includes calculating the prediction value of the prediction target by inputting, to the prediction model, a factor value for prediction about each of one or more factors including the factor selected in the factor selection processing and not including the factor excluded in the factor filtering processing.