Related references
Note: Only part of the references are listed.
Article
Business, Finance
Wan -Lin Yan
Summary: Stock index futures is a useful tool for stock investors to manage risk. This paper proposes a hybrid model combining AdaBoost feature selection and deep learning for predicting stock index futures prices. The results show that the model outperforms other popular prediction models such as random forest, multi-layer perception, gated recurrent unit, deep belief network and stacked denoising autoencoder.
NORTH AMERICAN JOURNAL OF ECONOMICS AND FINANCE
(2023)
Article
Energy & Fuels
Hasnain Iftikhar et al.
Summary: This study provides a comprehensive analysis of monthly electricity consumption forecasting by comparing decomposition techniques and time series models. The proposed framework, applied to Pakistan's electricity consumption data, shows highly accurate and efficient results. The proposed decomposition methods outperform benchmark ones and the final forecasting models produce the lowest mean error.
Article
Thermodynamics
Manuel Monge et al.
Summary: This study examines the behavior of the West Texas Intermediate crude oil price and the Baltic Dry Index under the assumption of geopolitical risks, using monthly data from January 1985 until May 2021. The results suggest that geopolitical risks and the Baltic Dry Index series will return to their original trends in the event of an exogenous shock, while the behavior of the West Texas Intermediate crude oil price is different. The findings are further supported by analyzing the long-term relationship of the time series using the Fractional Cointegration Vector AutoRegressive approach. Finally, Bai and Perron (2003) and wavelet transform approaches are used to detect breaks in maritime transport prices and crude oil prices caused by geopolitical risks.
Article
Thermodynamics
Xiaoping Xiong et al.
Summary: This paper proposes a new hybrid forecasting framework to improve the forecasting accuracy of day-ahead electricity prices. The proposed model consists of three valuable strategies: an adaptive copula-based feature selection algorithm, a new method of signal decomposition technique based on decomposition denoising strategy, and a Bayesian optimization and hyperband optimized long short-term memory model. Cross-validation using five datasets from the PJM electricity market shows that the proposed hybrid algorithm is more effective and practical for day-ahead electricity price forecasting.
Article
Engineering, Electrical & Electronic
Yong Qiang Tan et al.
Summary: Day-ahead electricity price forecasting is crucial in liberalized and deregulated electricity markets, providing references for bidding strategies, energy trading, and risk management. However, due to uncertain factors, electricity prices exhibit nonlinearity, randomness, and volatility, posing challenges for accurate forecasting. This study proposes a novel hybrid deep learning model, combined with improved decomposition and prediction techniques, which effectively improves prediction accuracy and stability in the Australian electricity market, showcasing outstanding performance in predicting price spikes. Additionally, the proposed model saves training time for neural networks due to its faster convergence speed, making it a valuable technology-based reference for electricity market participants.
ELECTRIC POWER SYSTEMS RESEARCH
(2023)
Article
Automation & Control Systems
Quande Qin et al.
Summary: Recent research has shown that introducing online data can significantly improve forecasting ability. In this study, several popular single-model machine learning methods and a stacking multiple-model ensemble learning strategy are used with online data from Google Trends to forecast crude oil prices. The study investigates the effect of Google Trends on crude oil prices and compares multiple-model methods with single-model machine learning methods. The results indicate that introducing Google Trends improves forecasting performance and multiple-model methods outperform single-model machine learning methods in terms of prediction accuracy.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Energy & Fuels
Leonard Tschora et al.
Summary: This article investigates the capabilities of different machine learning techniques in accurately predicting electricity prices. By considering previously unused predictive features such as price histories of neighboring countries, the current state-of-the-art approaches are extended. It is shown that these features significantly improve the quality of forecasts, even during periods of sudden changes. Analyzing the contribution of different features in model prediction using Shap values further sheds light on how models make their predictions and builds user confidence.
Article
Mathematics, Interdisciplinary Applications
Oznur Oztunc Kaymak et al.
Summary: This paper discusses the impact of global economic recession and the COVID-19 pandemic on oil prices, and aims to improve a model for more accurate predictions. The study finds that this model outperforms other models and artificial neural networks in forecasting crude oil prices.
CHAOS SOLITONS & FRACTALS
(2022)
Article
Automation & Control Systems
Yuxiong Li et al.
Summary: Remaining useful life prediction is crucial for preventing equipment malfunctions and reducing maintenance costs. Machine learning algorithms have gained popularity in this field due to their flexibility and convenience. However, they require large amounts of data and their prediction performance relies heavily on hyper-parameter selection. To address these limitations, a novel method based on multi-support vector regression fusion is proposed for remaining useful life prediction in small sample cases.
Article
Engineering, Multidisciplinary
Ibham Veza et al.
Summary: The conventional experimental methods for determining biomass heating value are laborious and costly. This study develops an ANN model trained with 11 different algorithms to predict the HHV of biomass samples. The results show that the ANN trained with Levenberg-Marquardt algorithm achieves the highest accuracy.
RESULTS IN ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Yajiao Tang et al.
Summary: This paper provides a timely review of the adoption of machine learning in financial time series (FTS) forecasting. The progress and application of machine learning methods in FTS forecasting models are systematically summarized, providing a relevant reference for researchers and practitioners. The paper identifies commonly used models and discusses their merits and demerits, as well as the limitations and future research directions of machine learning models in FTS forecasting.
Article
Green & Sustainable Science & Technology
Casper Boongaling Agaton
Summary: This study analyzes the dynamics of investment decisions in energy transition under uncertainties in geopolitical risks, using real options valuation to evaluate the flexibility in postponing energy transition investments. The study finds the feasibility of energy transition projects and the advantage of delaying implementation, emphasizing the importance of considering geopolitical factors in energy transition projects and providing policy guidance.
Article
Akshit Kurani et al.
Annals of Data Science
(2021)
Article
Thermodynamics
Pingping Xiong et al.
Summary: A new fractional-order accumulation-based incomplete gamma grey forecasting model is proposed in this paper, which combines dynamic nonlinear action-based incomplete gamma functions with fractional-order accumulation, fully considers the role of new information, and demonstrates good prediction performance for natural gas consumption in the Asia-Pacific region.
Article
Computer Science, Artificial Intelligence
Han-Min Kim et al.
Summary: With the growing interest in cryptocurrency, studies of cryptocurrency price predictions are actively conducted in various academic disciplines. Blockchain-specific information plays a key role in forecasting cryptocurrency values, with macro-economy factors and information from other publicly available coins also influencing prices.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Metallurgy & Metallurgical Engineering
Jing Guo Ding et al.
ISIJ INTERNATIONAL
(2021)
Article
Environmental Sciences
Floris Calkoen et al.
Summary: The study evaluates eight different time-series forecasting methods for predicting future shorelines, finding that traditional forecast methods and some evaluated probabilistic Machine Learning (ML) algorithms outperform Ordinary Least Squares (OLS) for the majority of the sites. Machine Learning algorithms show promise for more efficient predictions at a global scale, potentially outperforming both OLS regression and traditional time-series forecast methods in the future.
Article
Chemistry, Analytical
Taewon Moon et al.
Summary: Greenhouses require accurate and reliable data for resource use efficiency, and the study found that the U-Net architecture can be used to impute missing data, accurately learning and simulating changing patterns in the greenhouse environment, which can increase crop productivity and quality.
Article
Energy & Fuels
Dimitrios Mouchtaris et al.
Summary: This study attempted to forecast natural gas spot prices at different time frames using various machine learning methods, and found that the bagged trees and linear support vector machine models outperformed other models in forecasting performance.
Article
Green & Sustainable Science & Technology
So-Won Choi et al.
Summary: This study developed the Engineering Machine-learning Automation Platform (EMAP) using machine-learning technology, to analyze big data generated in various stages of EPC projects. By predicting contractor risks and supporting decisions based on data from bidding, engineering, construction, and OM stages, the study aimed to enhance risk management and decision-making capabilities.
Proceedings Paper
Computer Science, Artificial Intelligence
Afan Galih Salman et al.
Summary: This article proposes an Auto-regressive Integrated Moving Average (ARIMA) model to forecast better visibility for the variant value of parameters p, d, q using the grid technique. The experiment showed that ARIMA has the lowest MSE value of 0.00029 and a coefficient of variation value of 0.00315. The greater number of prediction data in the ARIMA model increases the MSE value.
5TH INTERNATIONAL CONFERENCE ON COMPUTER SCIENCE AND COMPUTATIONAL INTELLIGENCE 2020
(2021)
Article
Mathematics, Applied
Alex Sherstinsky
PHYSICA D-NONLINEAR PHENOMENA
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Nalini Gupta et al.
11TH INTERNATIONAL CONFERENCE ON AMBIENT SYSTEMS, NETWORKS AND TECHNOLOGIES (ANT) / THE 3RD INTERNATIONAL CONFERENCE ON EMERGING DATA AND INDUSTRY 4.0 (EDI40) / AFFILIATED WORKSHOPS
(2020)
Article
Computer Science, Artificial Intelligence
Li Yang et al.
Article
Energy & Fuels
Moting Su et al.
Article
Physics, Multidisciplinary
Sibao Fu et al.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2019)
Article
Computer Science, Artificial Intelligence
Alaa Sagheer et al.
Article
Business, Finance
Jonathan Brogaard et al.
REVIEW OF FINANCIAL STUDIES
(2019)
Article
Computer Science, Artificial Intelligence
Tomoumi Takase et al.
Article
Computer Science, Artificial Intelligence
Jie Cai et al.
Article
Multidisciplinary Sciences
Spyros Makridakis et al.
Article
Geosciences, Multidisciplinary
Shuang Gao et al.
GEOSCIENCE FRONTIERS
(2017)
Article
Computer Science, Artificial Intelligence
Salvador Garcia et al.
KNOWLEDGE-BASED SYSTEMS
(2016)
Article
Transportation
Veerachai Gosasang et al.
ASIAN JOURNAL OF SHIPPING AND LOGISTICS
(2011)
Article
Economics
Xun Zhang et al.
Article
Computer Science, Cybernetics
Alireza Farhangfar et al.
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART A-SYSTEMS AND HUMANS
(2007)
Review
Health Care Sciences & Services
A. Rogier T. Donders et al.
JOURNAL OF CLINICAL EPIDEMIOLOGY
(2006)