Automatic Spatial Regression Model Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Spatial regression models face uncertainty due to the selection of inaccurate models and spatial weights matrices, leading to inconsistent parameter estimates and flawed inferences, which can result in poor decision-making.
Innovation Solution
A method for selecting a trained spatial regression model by iterating through multiple models and spatial weights matrices, using a fit criterion to determine the best model and weights matrix combination, which is computationally efficient and suitable for big data applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple spatial regression models and spatial weights matrices are evaluated to ensure accurate model selection, then model accuracy and reliability are improved, but computational time and complexity increase
Solution Approach 1:
The patent applies preliminary action by pre-defining a comprehensive set of candidate spatial regression models and spatial weights matrices before the actual model selection process. This allows the system to systematically evaluate multiple pre-prepared options rather than generating them during analysis, reducing computational overhead while maintaining thoroughness in model comparison and selection accuracy.
Solution Approach 2:
The patent segments the model selection process into distinct evaluation stages, where each candidate model and weights matrix combination is assessed independently using a fit criterion. This segmentation allows for systematic comparison of multiple models without requiring exhaustive simultaneous computation, thereby managing computational complexity while ensuring reliable model selection through structured evaluation.
2Measurement precision
If a comprehensive set of spatial regression models is evaluated to identify the best model, then model selection accuracy is improved, but device complexity and computational resources required increase
Solution Approach 1:
The patent employs parameter changes by varying the fit criterion parameters across different candidate models and spatial weights matrices. By systematically changing evaluation parameters such as model type specifications and weights matrix configurations, the system achieves comprehensive model assessment without requiring complex custom evaluation logic for each model, thereby maintaining measurement precision while managing system complexity through parameterized evaluation.
3Ease of manufacture
If an inaccurate spatial weights matrix is used, then computational simplicity is maintained, but neighbor relationship accuracy deteriorates leading to flawed inferences
Solution Approach 1:
The patent applies self-service by enabling the spatial weights matrix to automatically adapt to the specific characteristics of the spatial data being analyzed. Rather than requiring manual specification or simplification of the weights matrix for computational ease, the system evaluates multiple candidate weights matrices and selects the one that best fits the data, allowing the analysis process to self-optimize the neighbor relationship representation while maintaining computational tractability through automated selection.
Data Source
AI summary
A computing device selects a trained spatial regression model. A spatial weights matrix defined for observation vectors is selected, where each element of the spatial weights matrix indicates an amount of influence between respective pairs of observation vectors. Each observation vector is spatially referenced. A spatial regression model is selected from spatial regression models, initialized, and trained using the observation vectors and the spatial weights matrix to fit a response variable using regressor variables. Each observation vector includes a response value for the response variable and a regressor value for each regressor variable of the regressor variables. A fit criterion value is computed for the spatial regression model and the spatial regression model selection, initialization, and training are repeated until each spatial regression model is selected. A best spatial regression model is selected and output as the spatial regression model having an extremum value of the fit criterion value.