Road Condition Prediction Using Bayesian Classification

The patent transforms the prediction approach by changing from a single parameter (travelling speed) to multiple parameters including travelling speed, road segment characteristics, time features, and location information. This parameter expansion resolves the contradiction by incorporating more comprehensive data dimensions while maintaining computational feasibility through structured feature engineering and probability-based classification.

The patent creates a composite prediction model that integrates multiple independent factors (speed, road characteristics, temporal features, spatial location) into a unified road condition status prediction system. This composite approach combines diverse data sources and feature types to achieve more accurate and reliable predictions compared to single-factor methods.

The patent segments the complex prediction problem into independent feature dimensions (travelling speed features, road segment characteristics, time features, location information). Each segment is processed and evaluated separately, then integrated through probability multiplication. This segmentation reduces computational complexity by avoiding complex interactions between features while maintaining comprehensive analysis.

The patent introduces probability values as intermediary representations for each classification event. Instead of directly complexly interacting multiple factors, the system converts each factor into a probability score that serves as an intermediary, then combines these intermediaries through multiplication to achieve the final prediction. This intermediary approach simplifies the integration of multiple factors.