Development of Knowledge Discovery Systems by using the Hierarchical Bayesian Time Series Models

使用分层贝叶斯时间序列模型开发知识发现系统

• 批准号: 12558023
• 负责人: HIGUCHI Tomoyuki
• 金额: $ 6.08万
• 依托单位: The Institute of Statistical Mathematics
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B)
• 财政年份: 2000
• 资助国家: 日本
• 起止时间: 2000 至 2003
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-12558023/
• 关键词: Hierarchical Bayesian model; Self-organizing; Generalized state space model; Particle filter; Hyperparameter; Count data; Monte Carlo method; Model averaging; モデル結合; 状態空間表現; 自己組織型モデル; モンテカルロフィルタ; ベイズ推定; 並列計算; PCクラスター; 物理乱数

In this project, we dealt with a removal of the artificial noises that is a stumbling block in the effort to perform an automatic procedure for knowledge discovery from a large-scale time series data. More specifically, we focused on the problems to exclude a rapid change in a trend (background mean) due to changes in sensitivity of the observation instruments, and to identify an outlier. The self-organizing state space model which belongs to the hierarchical Bayesian model has been employed to solve these problems. It is capable of estimating the trend component even if the noise component in a time series shows a time-dependent structure ; ex., its variance depends on time. The program that we developed allows us to detect the tune-dependent mean structure automatically for a large-scale time series datasets. We hope this program will open a door for us to re-analyze huge accumulated dataset that has not been examined in detail owing to an apparent signal contamination by various noises. We have already post this program with an explanation for usage on Web.http://tswww.ism.ac.jp/higuchi/index e/Soft/index.htm

在这个项目中，我们致力于消除人工噪声，这是从大规模时间序列数据中执行自动知识发现过程的绊脚石。更具体地说，我们关注的问题是排除由于观测仪器灵敏度变化而导致的趋势（背景平均值）的快速变化，并识别异常值。属于分层贝叶斯模型的自组织状态空间模型被用来解决这些问题。即使时间序列中的噪声成分呈现出与时间相关的结构，它也能够估计趋势成分；例如，它的方差取决于时间。我们开发的程序允许我们自动检测大规模时间序列数据集的依赖于调的平均结构。我们希望这个程序将为我们打开一扇门，让我们重新分析大量积累的数据集，这些数据集由于各种噪声的明显信号污染而尚未经过详细检查。我们已经在 Web 上发布了该程序及其使用说明。http://tswww.ism.ac.jp/higuchi/index e/Soft/index.htm

项目成果

T.Ikoma, N.Ichimura, T Higuchi, H.Maeda: "Particle Filter Based Method for Maneuvering Target Tracking"IEEE International Workshop on Intelligent Signal Processing. 3-8 (2001)

T.Ikoma、N.Ichimura、T Higuchi、H.Maeda：“基于粒子滤波器的机动目标跟踪方法”IEEE 国际智能信号处理研讨会。

G.Ueno,N.Nakamura,T.Higuchi,T,Tsuchiya,S.Machida,and T.Araki: "Application of Multivariate Maxwellian Mixture Model to Plasma Velocity Distribution Function"The proceedings of The Second International Conference on Discovery Science. AI series 1967. 197-2

G.Ueno、N.Nakamura、T.Higuchi、T、Tsuchiya、S.Machida 和 T.Araki：“多元麦克斯韦混合模型在等离子体速度分布函数中的应用”第二届国际发现科学会议论文集。

M.Kamiyama, T.Higuchi: "Non-linear filtering approach to an adjustment of non-uniform sampling locations in spatial datasets"Proceedings of 2003 IEEE Workshop on Statistical Signal Processing. 181-184 (2003)

M.Kamiyama、T.Higuchi：“调整空间数据集中非均匀采样位置的非线性过滤方法”2003 年 IEEE 统计信号处理研讨会论文集。

H.Nagao, T.Iyemori, T.Higuchi, T.Araki: "Lower Mantle Conductivity Anomalies Estimated from Geomagnetic Jerks"Journal of Geophysical Research-Solid Earth. (印刷中).

H.Nagao、T.Iyemori、T.Higuchi、T.Araki：“根据地磁急动估计的下地幔电导率异常”地球物理研究杂志-固体地球（正在出版）。

G.Kitagawa, T.Higuchi, F.N.Kondo: "Smoothness Prior Approach to Explore Mean Structure in Large Time Series"Theoretical Computer Science. (印刷中). (2002)

G.Kitakawa、T.Higuchi、F.N.Kondo：“探索大型时间序列中的平均结构的平滑先验方法”理论计算机科学（2002 年）。