Attention based hybrid parametric and neural network models for non-stationary time series prediction

Article Energy & Fuels

Forecasting underheating in dwellings to detect excess winter mortality risks using time series models

Ahmed I. Ahmed et al.

Summary: The study developed a novel recursive time series model combining AutoRegressive with eXogenous inputs to provide reliable predictions of wintertime internal temperatures in homes, achieving Mean Absolute Errors below 1.38 degrees C for 1, 3, and 6-hour forecasts. The research demonstrated the superiority of this model in forecasting, showing potential as part of a Winter Early Warning Response System for identifying homes at risk of cold-related health impacts.

APPLIED ENERGY (2021)

Add to Collection

Article Physics, Multidisciplinary

Forecasting price of financial market crash via a new nonlinear potential GARCH model

Dun-Zhong Xing et al.

Summary: Analyzing the underlying structure and forecasting methods of financial market crashes is crucial for market stability. The study introduces a new nonlinear potential GARCH model, combining likelihood function estimations to analyze the effects of potential functions and fluctuations on prices, showing significant changes in the potential well of the underlying nonlinear potential function during financial price crashes.

PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

A HYBRID MODEL INTEGRATING LSTM AND GARCH FOR BITCOIN PRICE PREDICTION

Zidi Gao et al.

Summary: The study suggests that combining parametric models like GARCH with deep neural networks can lead to better results in cryptocurrency price forecasting, especially in the case of high volatility data or limited data sequences.

2021 IEEE 31ST INTERNATIONAL WORKSHOP ON MACHINE LEARNING FOR SIGNAL PROCESSING (MLSP) (2021)

Add to Collection

Article Computer Science, Information Systems

Attention-Based Bi-Directional Long-Short Term Memory Network for Earthquake Prediction

Md. Hasan Al Banna et al.

Summary: This paper proposes an earthquake occurrence and location prediction model using LSTM architecture and attention mechanism, achieving successful predictions of earthquake events and epicenters in Bangladesh.

IEEE ACCESS (2021)

Add to Collection

Review Computer Science, Information Systems