Incremental learning based on the nearest neighbor classifier
Sunfu Liu, Qing Ye, Xiang Liu · 2015
In many practical applications, it is difficult to obtain the complete sample set and the categories may be variable with time. This paper introduces an online incremental learning algorithm based on the nearest neighbor algorithm to increase model or category during the process of recognition. Calculating the matching degree between new input sample and model samples, the algorithm finds the best and second best matching degrees and compares them with the threshold. The comparative results decide whether it increases samples' amount or adds a new category to realize incremental learning. The algorithm is applied to segment data set and experiment of vehicle type recognition. The experiments prove the algorithm efficient. Also, more experiments are conducted to analyze and verify how standard sample amount and matching degree threshold affect the final results.