Time series prediction for EMS with machine learning

A probabilistic learning technique, known as gated mixture of experts (MEs), is made more adaptive by employing a customized genetic algorithm based on the concepts of hierarchical mixed encoding and hybrid training. The objective of such effort is to promote the automatic design (i.e., structural configuration and parameter calibration) of whole gated ME instances more capable to cope with the intricacies of some difficult machine learning problems whose statistical properties are time-variant. In this chapter, we outline the main steps behind such novel hybrid intelligent system, focusing on its application to the nontrivial task of nonlinear time-series forecasting. Experiment results are reported with respect to three benchmarking time-series problems, and confirmed our expectation that the new integrated approach is capable to outperform, both in terms of accuracy and generalization, other conventional approaches, such as single neural networks and non-adaptive, handcrafted gated MEs.

Download Full-text

Time series prediction using machine learning: a case of Bitcoin returns

Studies in Economics and Finance ◽

10.1108/sef-06-2021-0217 ◽

2021 ◽

Vol ahead-of-print (ahead-of-print) ◽

Author(s):

Irfan Haider Shakri

Keyword(s):

Machine Learning ◽

Time Series ◽

Time Series Data ◽

Time Series Prediction ◽

Predictive Ability ◽

Absolute Error ◽

Data Driven ◽

Series Data ◽

Policy Uncertainty ◽

Content Type

Purpose The purpose of this study is to compare five data-driven-based ML techniques to predict the time series data of Bitcoin returns, namely, alternating model tree, random forest (RF), multiple linear regression, multi-layer perceptron regression and M5 Tree algorithms. Design/methodology/approach The data used to forecast time series data of Bitcoin returns ranges from 8 July 2010 to 30 Aug 2020. This study used several predictors to predict bitcoin returns including economic policy uncertainty, equity market volatility index, S&P returns, USD/EURO exchange rates, oil and gold prices, volatilities and returns. Five statistical indexes, namely, correlation coefficient, mean absolute error, root mean square error, relative absolute error and root relative squared error are determined. The results of these metrices are used to develop colour intensity ranking. Findings Among the machine learning (ML) techniques used in this study, RF models has shown superior predictive ability for estimating the Bitcoin returns. Originality/value This study is first of its kind to use and compare ML models in the prediction of Bitcoins. More studies can be carried out by using further cryptocurrencies and other ML data-driven models in future.

Download Full-text

Evaluation of statistical and machine learning models for time series prediction: Identifying the state-of-the-art and the best conditions for the use of each model

Information Sciences ◽

10.1016/j.ins.2019.01.076 ◽

2019 ◽

Vol 484 ◽

pp. 302-337 ◽

Cited By ~ 17

Author(s):

Antonio Rafael Sabino Parmezan ◽

Vinicius M.A. Souza ◽

Gustavo E.A.P.A. Batista

Keyword(s):

Machine Learning ◽

Time Series ◽

State Of The Art ◽

Time Series Prediction ◽

The State ◽

Learning Models ◽

Machine Learning Models

Download Full-text

Combining Measures of Signal Complexity and Machine Learning for Time Series Analyis: A Review

Entropy ◽

10.3390/e23121672 ◽

2021 ◽

Vol 23 (12) ◽

pp. 1672

Author(s):

Sebastian Raubitzek ◽

Thomas Neubauer

Keyword(s):

Machine Learning ◽

Time Series ◽

Deep Learning ◽

Time Series Prediction ◽

Long Term Memory ◽

Learning Approaches ◽

Entropy Measures ◽

Combine Machine ◽

Signal Complexity

Measures of signal complexity, such as the Hurst exponent, the fractal dimension, and the Spectrum of Lyapunov exponents, are used in time series analysis to give estimates on persistency, anti-persistency, fluctuations and predictability of the data under study. They have proven beneficial when doing time series prediction using machine and deep learning and tell what features may be relevant for predicting time-series and establishing complexity features. Further, the performance of machine learning approaches can be improved, taking into account the complexity of the data under study, e.g., adapting the employed algorithm to the inherent long-term memory of the data. In this article, we provide a review of complexity and entropy measures in combination with machine learning approaches. We give a comprehensive review of relevant publications, suggesting the use of fractal or complexity-measure concepts to improve existing machine or deep learning approaches. Additionally, we evaluate applications of these concepts and examine if they can be helpful in predicting and analyzing time series using machine and deep learning. Finally, we give a list of a total of six ways to combine machine learning and measures of signal complexity as found in the literature.

Download Full-text