Dataset Information

Autoreservoir computing for multistep ahead prediction based on the spatiotemporal information transformation.

ABSTRACT: We develop an auto-reservoir computing framework, Auto-Reservoir Neural Network (ARNN), to efficiently and accurately make multi-step-ahead predictions based on a short-term high-dimensional time series. Different from traditional reservoir computing whose reservoir is an external dynamical system irrelevant to the target system, ARNN directly transforms the observed high-dimensional dynamics as its reservoir, which maps the high-dimensional/spatial data to the future temporal values of a target variable based on our spatiotemporal information (STI) transformation. Thus, the multi-step prediction of the target variable is achieved in an accurate and computationally efficient manner. ARNN is successfully applied to both representative models and real-world datasets, all of which show satisfactory performance in the multi-step-ahead prediction, even when the data are perturbed by noise and when the system is time-varying. Actually, such ARNN transformation equivalently expands the sample size and thus has great potential in practical applications in artificial intelligence and machine learning.

SUBMITTER: Chen P

PROVIDER: S-EPMC7486927 | biostudies-literature | 2020 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Autoreservoir computing for multistep ahead prediction based on the spatiotemporal information transformation.

Chen Pei P Liu Rui R Aihara Kazuyuki K Chen Luonan L

Nature communications 20200911 1

We develop an auto-reservoir computing framework, Auto-Reservoir Neural Network (ARNN), to efficiently and accurately make multi-step-ahead predictions based on a short-term high-dimensional time series. Different from traditional reservoir computing whose reservoir is an external dynamical system irrelevant to the target system, ARNN directly transforms the observed high-dimensional dynamics as its reservoir, which maps the high-dimensional/spatial data to the future temporal values of a target ...[more]

PMID: 32917894

Similar Datasets

Project description:User-generated content (UGC) is an important data source for tourism GIScience research. However, no effective approach exists for identifying hidden spatiotemporal patterns within multi-scale unstructured UGC. Therefore, we developed an algorithm to measure the tourist destination popularity (TDP) based on a multi-spatiotemporal text granular computing model, called TDPMTGC. To accurately granulate the spatial and temporal information of tourism text, tourism text data granules are used to represent landscape objects. These granules are unified objects that possess multiple attributes, such as spatial and temporal dimensions. The multi-spatiotemporal scales are characterized by the multi-hierarchical structure of granular computing, and transformations of granular layers and data granule size are achieved by scale selection in the spatial and temporal dimensions. Therefore, all scales between the spatial and temporal dimension are related, which allows for the comparability of the data granules of all spatial-spatial, temporal-temporal and spatial-temporal layers. This approach achieves a quantitative description and comparison of the popularity value of granules between adjacent scales and cross-scales. Therefore, the TDP with multi-spatiotemporal scales can be deduced and calculated in a systematic framework. We first introduce the conceptual framework of TDPMTGC to construct a quantitative measurement model of TDP at multi-spatiotemporal scales. Then, we present a dataset construction approach to support multi-spatiotemporal scale granular reorganization. Finally, TDPMTGC is derived to describe both the TDP at a single spatial or temporal scale and the patterns and processes of the TDP at multi-spatiotemporal scales. A case study from Jiuzhaigou shows that the TDP derived using TDPMTGC is consistent with the conclusions of existing studies. More importantly, TDPMTGC provides additional detailed characteristics, such as the contributions of different scenic spots in a tourist route or scenic area, the monthly anomalies and daily contributions of TDP in a specific year, the distinct weakening of tourist route scale in tourist cognition, and the daily variations of TDP during in-season and off-season times. This is the first time that a granular computing model has been introduced to tourism GIScience that provides a feasible scheme for reorganizing large-scale unstructured text and constructing public spatiotemporal UGC tourism datasets. TDPMTGC constitutes a new approach for exploring tourist behaviors and the driving mechanisms of tourism patterns and processes.

Dataset Information

Autoreservoir computing for multistep ahead prediction based on the spatiotemporal information transformation.

Publications

Autoreservoir computing for multistep ahead prediction based on the spatiotemporal information transformation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets