在决策边界理论中,决策边界是最佳分类器(optimal classifier),即能使归 类准确性最大化的装置。但人类在许多方面不如最佳分类器:(1)不知道每一类 别的每一样例的位置,(2)不知道知觉噪...
基于8个网页-相关网页
分类器组合的有效性问题以及最佳组合问题均需要解决。
The effectiveness of classifier combination and the problem of best combination both have to be solved.
相似度匹配方法的研究结果表明余弦距离分类器分类效果最佳。
Studying of similarity match methods, it shows that cosine distance is best for classification.
在实际应用中,一般的近邻分类器由于模式处理量过大,且难以在线和快速获得最佳近邻数等原因,而受到了限制。
In practice applications, generic near neighbor classifier was limited for the amount of pattern processing was very large and was difficult to get the best data quickly on line.
应用推荐