Spatiotemporal variability of Indian rainfall using multiscale entropy

Journal of Hydrology, Год журнала: 2020, Номер 587, С. 124916 - 124916

Опубликована: Апрель 3, 2020

Язык: Английский

---

A Labeling Method for Financial Time Series Prediction Based on Trends

Entropy, Год журнала: 2020, Номер 22(10), С. 1162 - 1162

Опубликована: Окт. 15, 2020

Time series prediction has been widely applied to the finance industry in applications such as stock market price and commodity forecasting. Machine learning methods have used financial time recent years. How label data determine accuracy of machine models subsequently final investment returns is a hot topic. Existing labeling mainly by comparing current with those short period future. However, are typically non-linear obvious short-term randomness. Therefore, these not captured continuous trend features data, leading difference between their results real trends. In this paper, new method called “continuous labeling” proposed address above problem. feature preprocessing stage, paper that can avoid problem look-ahead bias traditional standardization or normalization processes. Then, detailed logical explanation was given, definition also an automatic algorithm given extract data. Experiments on Shanghai Composite Index Shenzhen Component some stocks China showed our much better state-of-the-art terms classification other evaluation metrics. The proved deep LSTM GRU more suitable for dealing

Язык: Английский

---

Quantile-based Bayesian Model Averaging approach towards merging of precipitation products

Journal of Hydrology, Год журнала: 2021, Номер 604, С. 127206 - 127206

Опубликована: Ноя. 25, 2021

Язык: Английский

---

Studying the impact of fluctuations, spikes and rare events in time series through a wavelet entropy predictability measure

Physica A Statistical Mechanics and its Applications, Год журнала: 2024, Номер 641, С. 129720 - 129720

Опубликована: Март 30, 2024

Язык: Английский

---

Accounting for temporal variability for improved precipitation regionalization based on self-organizing map coupled with information theory

Journal of Hydrology, Год журнала: 2020, Номер 590, С. 125236 - 125236

Опубликована: Июль 4, 2020

Язык: Английский

---

Forecasting of extreme flood events using different satellite precipitation products and wavelet-based machine learning methods

Chaos An Interdisciplinary Journal of Nonlinear Science, Год журнала: 2020, Номер 30(6)

Опубликована: Июнь 1, 2020

An accurate and timely forecast of extreme events can mitigate negative impacts enhance preparedness. Real-time forecasting flood with longer lead times is difficult for regions sparse rain gauges, in such situations, satellite precipitation could be a better alternative. Machine learning methods have shown promising results minimum variables indicating the underlying nonlinear complex hydrologic system. Integration machine event motivates us to develop reliable models that are simple, accurate, applicable data scare regions. In this study, we method using product wavelet-based models. We test proposed approach flood-prone Vamsadhara river basin, India. The validation show has potential comparison other benchmark

Язык: Английский

---

Critical slowing down indicators

EPL (Europhysics Letters), Год журнала: 2020, Номер 132(1), С. 18001 - 18001

Опубликована: Окт. 1, 2020

Abstract Critical slowing down is considered to be an important indicator for predicting critical transitions in dynamical systems. Researchers have used it prolifically the fields of ecology, biology, sociology, and finance. When a system approaches transition or tipping point, returns more slowly its stable attractor under small perturbations. The return time state can thus as index, which shows whether change near not. Based on this phenomenon, many methods been proposed determine points, especially biological social systems, example, related epidemic spreading, cardiac arrhythmias, even population collapse. In perspective, we briefly review past research dedicated indicators associated outline promising directions future research.

Язык: Английский

---

One or two frequencies? The Iterative Filtering answers

Applied Mathematics and Computation, Год журнала: 2023, Номер 462, С. 128322 - 128322

Опубликована: Сен. 13, 2023

The Iterative Filtering method is a technique aimed at the decomposition of non-stationary and non-linear signals into simple oscillatory components. This method, proposed decade ago as an alternative to Empirical Mode Decomposition, has been used extensively in many applied fields research studied, from mathematical point view, several papers published last few years. However, even if its convergence stability are now established both continuous discrete setting, it still open problem understand up what extent this approach can separate two close-by frequencies contained signal. In paper, first we recall previously discovered theoretical results about Filtering. Afterward, prove new theorems regarding ability separating nearby case continuously sampled signals. Among them, theorem which allows construct filters captures, machine precision, specific frequency. We run numerical tests confirm our findings compare performance with one Decomposition Synchrosqueezing methods. All presented under investigation addressing fundamental "one or frequencies" question.

Язык: Английский

---

Combining Seasonal and Trend Decomposition Using LOESS With a Gated Recurrent Unit for Climate Time Series Forecasting

IEEE Access, Год журнала: 2024, Номер 12, С. 85275 - 85290

Опубликована: Янв. 1, 2024

Язык: Английский

---

Wavelet analysis and energy-based measures for oil-food price relationship as a footprint of financialisation effect

Resources Policy, Год журнала: 2022, Номер 77, С. 102692 - 102692

Опубликована: Апрель 11, 2022

Язык: Английский

---