摘要
提出了一种对商品不同特征值缩放的方法,改进了KNN价格预测模型,并对改进后的预测结果做了研究分析。分析得出:对基于KNN算法的价格预测模型的改进是有效的,提高了预测准确度。
In the KNN price forecasting model,when all the characteristics of goods within the scope not in the same range,the prediction accuracy will be greatly reduced.The improvement of the KNN price forecasting model is the main contents of this article,presented the scale means for the value of different characteristics of goods,improved KNN price forecasting model,and analyzed the results of im-proved forecasting model.Experimental results show that: the improved KNN price forecasting model is effective to improve the prediction accuracy.
出处
《软件导刊》
2010年第10期84-86,共3页
Software Guide
基金
国家科技支撑计划项目(2007BAD33B05)
关键词
KNN算法
价格预测
特征缩放
预测模型改进
KNN Algorithm
Price Forecast
characteristic scale
Improved Forecast Model