Applied multivariate statistical analysis
Applied multivariate statistical analysis
Data Mining: Introductory and Advanced Topics
Data Mining: Introductory and Advanced Topics
Time Series Analysis: Forecasting and Control
Time Series Analysis: Forecasting and Control
Neural-network classifiers for recognizing totally unconstrained handwritten numerals
IEEE Transactions on Neural Networks
A comparison between neural-network forecasting techniques-case study: river flow forecasting
IEEE Transactions on Neural Networks
| Hi-index | 0.00 |
The importance of forecasting cannot be overemphasized in modern environment surveillance applications, including flood control, rainfall analysis, pollution study, nuclear leakage prevention and so on. That is why we proposed STIFF (SpatioTemporal Integrated Forecasting Framework) in previous work [11], trying to answer such a challenging problem of doing forecasting in natural environment with both spatial and temporal characteristics involved. However, despite its promising performance on univariate-based data, STIFF is not sophisticated enough for more complicated environmental data derived from multiple correlated variables. Therefore in this paper we add multivariate analysis to the solution, take the correlation between different variables into account and further extend STIFF to address spatiotemporal forecasting involving multiple variables. Our experiments show that this introduction and integration of multivariate correlation not only has a more reasonable and rational interpretation of the data itself, but also produces a higher accuracy and slightly more balanced behavior.