论文标题
个性化食品套件分布的聚类技术的比较分析
Comparative Analysis of Clustering Techniques for Personalized Food Kit Distribution
论文作者
论文摘要
由于大流行，喀拉拉邦政府增加了免费食品套件的供应频率，但是，这些项目是静态的，并不表明消费者的个人喜好。本文对通过基于联合分析的调查获得的真实数据集的缩放版本的各种聚类技术进行了比较分析。通过基于Centroid的方法（例如K平均值）进行聚类，并与SVD一起绘制结果，最后得出结论，即两者中的哪个更好。一旦制定了集群，也为每个集群决定商品。同样，基于特定的集群损失阈值，通过重新分配进一步增强了聚类。因此，最终获得了最有效的聚类技术，用于设计满足个人需求的食品套件。
The Government of Kerala had increased the frequency of supply of free food kits owing to the pandemic, however, these items were static and not indicative of the personal preferences of the consumers. This paper conducts a comparative analysis of various clustering techniques on a scaled-down version of a real-world dataset obtained through a conjoint analysis-based survey. Clustering carried out by centroid-based methods such as k means is analyzed and the results are plotted along with SVD, and finally, a conclusion is reached as to which among the two is better. Once the clusters have been formulated, commodities are also decided upon for each cluster. Also, clustering is further enhanced by reassignment, based on a specific cluster loss threshold. Thus, the most efficacious clustering technique for designing a food kit tailored to the needs of individuals is finally obtained.
