基于k-means与Apriori算法的食物营养成分分析

Analysis of Food Nutritional Components Based on k-means and Apriori Algorithm

营养作为人类生活的必要前提,大量患有某种疾病患者或由于工作职业原因对不同营养成分需求各不一致,发现不同食物种类营养成分及含量间的关系具有较强的应用价值。由于各类食物类别所含食物数量不同,针对Apriori算法通过支持度和置信度来衡量关联规则的特点,为克服各类食物数量不一致容易对挖掘结果产生不良影响,设计了一种通过k-means与Apriori算法对多种食物的营养成分及含量的挖掘与分析的方法。首先根据不同食物营养成分含量采用k-means聚类算法进行聚类,将食物数据集划分出了多个互不相交的"簇",再在各"簇"内通过Apriori算法实现食物营养成分含量之间的关联规则挖掘,其结果表明使用该方法经过聚类后的同一簇内食物营养成分关联程度明显优于直接在数据集中使用Apriori算法进行挖掘,为各类人群的合理膳食及饮食健康提供了重要的参考依据。

Nutrition is a necessary prerequisite for human＇s life. There are different food nutrition needs for a large number of ill patients or other persons due to occupational reasons. It has a strong application values if the relationship between different food types nutrients and content has found. But the number of foods contained in each food types is different. And the association rules is measured by support and confidence in Apriori algorithm. In order to overcome poor performance for mining results caused by different number of foods in each food categories. A method that both k-means and Apriori algorithm are adopted to realize the mining and analysis of food nutrition components was designed. First,k-means was used which has achieved cluster analysis according to nutrient content of different foods. And the food datasets were divided into a number of disjointed ＂clusters＂. Second,the association rule mining between food nutrient content was achieved by Apriori algorithm in each clusters. The results shows that the relationship between nutrient contents of each cluster＇s foods is better than direct mining it by Apriori obviously.An important reference was provided for human reasonable diet and diet health.