Nonlinear dynamics model for social popularity prediction based on multivariate chaotic time series
7
Citation
50
Reference
10
Related Paper
Citation Trend
Keywords:
Popularity
Popularity
Extreme Learning Machine
Sentiment Analysis
Categorical variable
Learning classifier system
Online machine learning
Cite
Citations (1)
A time-series analysis method of transient chaos is worked out which can also be applied to signals of laboratory experiments. The process is based on the construction of a long artificial time series obtained by gluing pieces of many transiently chaotic signals together. This artificial signal represents a long-time motion in the vicinity of the nonattracting chaotic set. Thus all of the well-known numerical methods developed for analyzing permanent chaotic behavior are applicable in a more convenient way than using many short separated time-series pieces. The method is illustrated and its validity is checked by the H\'enon map. The nonattracting strange set is reconstructed in the presence of both a periodic and a chaotic attractor, and quantitative characteristics such as dimensions and Lyapunov exponents are determined by means of time-delay embedding methods.
Transient (computer programming)
Chaos theory
Cite
Citations (34)
We discuss issues concerning the reconstruction of attractors from experimental chaotic time-series data using Takens's method of delays [in Proceedings of the Warwick Symposium, 1981, edited by D. A. Rand and L. S. Young (Springer, New York, 1981)]. The focus of this paper is on the selection of appropriate lag-time and embedding-dimension values with an emphasis on the relationship between those parameters and data-measurement considerations. We are particularly interested in the effect that low-pass filtering has on the appearance and measured properties of reconstructed attractors. Empirical results are presented using data measured from a laboratory fluidized bed and from data generated by integrating the Lorenz [J. Atmospheric Sci. 20, 130 (1963)] and Franceschini [Physica 6D, 285 (1983)] models of chaotic dynamic systems.
Lorenz system
Cite
Citations (15)
Lorenz system
Ant colony
Cite
Citations (63)
Given an irregular time series, an important issue is to determine whether it stems from a stochastic or a chaotic (i.e. deterministic with few degrees of freedom) system. This is generally achieved by studying the geometry of a reconstructed attractor, although it is known that some purely stochastic processes can be associated with low-dimension attractors. It is shown that an effective estimation of the number of degrees of freedom can be obtained better through a (local) independent component analysis.< >
Component (thermodynamics)
Cite
Citations (0)
A new approach to short-term load forecasting (STLF) in power systems is described in this paper. The method uses a chaotic time series and artificial neural network. The paper describes chaos time series analysis of daily power system peak loads. Nonlinear mapping of deterministic chaos is identified by multilayer perceptron (MLP). Using embedding dimension and delay time, an attractor in pseudo phase plane and an ANN model trained by this attractor are constructed. The proposed approach is demonstrated by an example.
Multilayer perceptron
Perceptron
Long-term prediction
Cite
Citations (21)
A method of short-term load forecasting based on chaotic time series and neural networks is presented in this paper.Firstly,attractors in phase spaces using chaotic theory is reconstructed.Secondly,the attractor's evolvement using BP neural networks is made,and the neural network's input data using Euclid distance is selected.The result analysis of the practical examples show that the proposed method is effective and feasible.
Cite
Citations (2)
Machine learning has been widely used in healthcare applications to approximate complex models, for clinical diagnosis, prognosis, and treatment. As deep learning has the outstanding ability to extract information from time series, its true capabilities on sparse, irregularly sampled, multivariate, and imbalanced physiological data are not yet fully explored. In this paper, we systematically examine the performance of machine learning models for the clinical prediction task based on the EHR, especially physiological time series. We choose Physionet 2019 challenge public dataset to predict Sepsis outcomes in ICU units. Ten baseline machine learning models are compared, including 3 deep learning methods and 7 non-deep learning methods, commonly used in the clinical prediction domain. Nine evaluation metrics with specific clinical implications are used to assess the performance of models. Besides, we sub-sample training dataset sizes and use learning curve fit to investigate the impact of the training dataset size on the performance of the machine learning models. We also propose the general pre-processing method for the physiology time-series data and use Dice Loss to deal with the dataset imbalanced problem. The results show that deep learning indeed outperforms non-deep learning, but with certain conditions: firstly, evaluating with some particular evaluation metrics (AUROC, AUPRC, Sensitivity, and FNR), but not others; secondly, the training dataset size is large enough (with an estimation of a magnitude of thousands).
Cite
Citations (1)
It is a piece of very meaningful task to identify whether the complex time series which has no apparent change in the rules is random or chaotic.The chaotic system usually can be identified by the existence of chaotic attractors.Around this characteristics,this paper discussed a number of chaotic time series identification methods.
Identification
Cite
Citations (0)
Before extracting the chaotic characteristic of the time series,it is needed to consider whether this time series have the chaos.If it is assumed that the experimental data is chaotic without examination in advance and the reconstruction phase space theory is used directly to extract the time series characteristics to build the model and the forecast,the result will be incredible.The chaotic system usually can be identified by the existence of chaotic attractors.Around this characteristics,this paper discussed the chaotic time series decision method.
Chaos theory
Cite
Citations (5)