从大数据中挖掘有价值的项

MINING VALUABLE ITEMS FROM BIG DATA

当我们收集大量具有相关的成本和一个隐含的值的项时,只有确定选择某个项以后,才能知道这个项隐含的值。假设给定一个累积成本的范围,如何从这些项中选出一个子集,使得其累积成本小于给定的范围,但是其隐含值的和最大。本文使用项之间的相似知识为先验知识,用核函数来找出满足条件的最小子集,并使用高斯过程来平衡估计项的值和选择值最大的项之间的能量消耗。另外,同时具有高使用性和多样性的子集也能被找到。

When we have a large collection of items,each with an associated cost and an inherent utility that is revealed only once we commit to selecting it. Suppose given a budget on the cumulative cost of the selected items,how can we pick a subset of maximal value? In this paper,an algorithm which utilizes prior knowledge about similarity between items,expressed as a kernel function. Gaussian process prediction was used to balance estimating the unknown value of items and selecting items of high value. Sets that simultaneously have high utility and are diverse are discovered,too.