A-Spatial-Missing-Value-Imputation-Method-for-Multi-view-Urban-Statistical-Data
一篇来自IJCAI’20的关于数据补全的work, 原文地址 Link
背景
当前城市数据存在残缺情况亟待解决。
作者针对确切的数据集 Australian Bureau of Statistics(ABS)提出了几点挑战:
- missing temporal information. 在此数据集下,每年残缺的部分比较规律,比如经常会出现某段时间周期性的缺失
- multi-view problem. 如果只从数据角度去修复可能会导致性能不佳,因此需要从multi-view,如经济、人口等等
- spatial correlation mining problem. 细粒度区域可能会发生明显的非线性变化,因此需要考虑一些区域相关性
To address all challenges, our proposed method is designed as a spatially related method which can only use spatial information to achieve a strong performance. In detail, the method integrates a spatial multi-kernel clustering method and an adaptive-weight non-negative matrix factorization (NMF) for solving the multi-view spatially related tasks.
文献综述
空间残缺数据补全
multi-view learning
多视图学习方法涉及不同视图的多样性,可以基于各种特征子集共同优化.
a matrix co-factorization based method (MVL-IV) 存在一些问题,当有不同视角残缺比例不同,coefficient matrix更倾向于向dense view学习 …
To the best of our knowledge, none of the above studies considered both spatial and multi-view problems. Hence, in this paper, we proposed an effective missing value imputation model for multi-view urban statistical data.
Method
…(⊙﹏⊙),写不下去了,multi-view learning看起来是个大坑,之后有时间再去接触一下叭。
给人的感觉好像是针对一个问题,我们有不同的 数据/view, e.g., income, population, economy, etc.
这个work看起来是针对spatial 和 mutli-view 进行了优化,提出了SMV-NMF算法,而且数据集具有特殊性,已经脱离spatial-temporal data imputation的范畴了。
留个小尾巴,之后接触相关知识再看
😬
A-Spatial-Missing-Value-Imputation-Method-for-Multi-view-Urban-Statistical-Data