SrVARM: State Regularized Vector Autoregressive Model for Joint Learning of Hidden State Transitions and State-Dependent Inter-Variable Dependencies from Multi-variate Time Series
SrVARM: State Regularized Vector Autoregressive Model for Joint Learning of Hidden State Transitions and State-Dependent Inter-Variable Dependencies from Multi-variate Time Series
复制标题
SrVARM：状态正则化向量自回归模型，用于联合学习多变量时间序列中的隐藏状态转换和状态相关变量间依赖性
DOI：
10.1145/3442381.3450116
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Honavar, Vasant G.
中科院分区：
文献类型：
--
作者：
Hsieh, Tsung-Yu;Sun, Yiwei;Tang, Xianfeng;Wang, Suhang;Honavar, Vasant G.
许多应用，例如，医疗保健、教育需要用于从高维时间序列数据构建预测模型的有效方法，其中变量之间的关系可能是复杂的并且随时间变化。在这样的设置中，底层系统在有限的隐藏状态集合之间经历了一系列不可观察的转换。此外，观测变量与其时间动态之间的关系可能取决于系统的隐藏状态。更复杂的是，来自不同个体的观测数据所隐含的隐藏状态序列可能不会相对于共同的参考系对齐。在此背景下，我们认为新的问题，共同学习的状态相关的变量间的关系，以及隐藏状态之间的过渡模式，从多变量的时间序列数据。为了解决这个问题，我们引入了状态正则化向量自回归模型（SrVARM），该模型将状态正则化递归神经网络与增强自回归模型相结合，以学习离散隐藏状态之间的动态转换，该增强自回归模型使用状态相关有向无环图（DAG）对每个状态中的变量间依赖关系进行建模。我们提出了一个有效的算法，利用最近推出的重新制定的组合问题，优化DAG结构相对于一个连续的优化问题的评分函数训练SrVARM。我们报告了模拟数据的广泛实验结果以及现实世界的基准测试，表明SrVARM在恢复未观察到的状态转换和发现变量之间的状态依赖关系方面优于最先进的基线。
Many applications, e.g., healthcare, education, call for effective methods methods for constructing predictive models from high dimensional time series data where the relationship between variables can be complex and vary over time. In such settings, the underlying system undergoes a sequence of unobserved transitions among a finite set of hidden states. Furthermore, the relationships between the observed variables and their temporal dynamics may depend on the hidden state of the system. To further complicate matters, the hidden state sequences underlying the observed data from different individuals may not be aligned relative to a common frame of reference. Against this background, we consider the novel problem of jointly learning the state-dependent inter-variable relationships as well as the pattern of transitions between hidden states from multi-variate time series data. To solve this problem, we introduce the State-Regularized Vector Autoregressive Model (SrVARM) which combines a state-regularized recurrent neural network to learn the dynamics of transitions between discrete hidden states with an augmented autoregressive model which models the inter-variable dependencies in each state using a state-dependent directed acyclic graph (DAG). We propose an efficient algorithm for training SrVARM by leveraging a recently introduced reformulation of the combinatorial problem of optimizing the DAG structure with respect to a scoring function into a continuous optimization problem. We report results of extensive experiments with simulated data as well as a real-world benchmark that show that SrVARM outperforms state-of-the-art baselines in recovering the unobserved state transitions and discovering the state-dependent relationships among variables.
登录
查看更多内容
影响因子：
3
作者：
Doberenz, Sigrun;Roth, Walton T.;Wollburg, Eileen;Maslowski, Nina I.;Kim, Sunyoung
通讯作者：
Kim, Sunyoung
通过深度神经网络进行可扩展因果图学习
DOI：
10.1145/3357384.3357864
发表时间：
2019
期刊：
Proceedings of the 28th ACM International Conference on Information and Knowledge Management
影响因子：
--
作者：
Chenxiao Xu;Hao Huang;Shinjae Yoo
通讯作者：
Shinjae Yoo
DOI：
10.1609/aaai.v35i10.17038
发表时间：
2020-05
期刊：
ArXiv
影响因子：
--
作者：
Junjie Liang;Yanting Wu;Dongkuan Xu;Vasant G Honavar
通讯作者：
Junjie Liang;Yanting Wu;Dongkuan Xu;Vasant G Honavar
DOI：
10.1109/embc.2018.8512928
发表时间：
2018
期刊：
2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC
影响因子：
--
作者：
Wickramasuriya, Dilranjan S.;Qi, Chaoxian;Faghih, Rose T.
通讯作者：
Faghih, Rose T.
DOI：
--
发表时间：
2019-05
期刊：
Proceedings of machine learning research
影响因子：
--
作者：
Biwei Huang;Kun Zhang;Mingming Gong;C. Glymour
通讯作者：
Biwei Huang;Kun Zhang;Mingming Gong;C. Glymour