Diego Barroso - Academia.edu (original) (raw)
Uploads
Papers by Diego Barroso
This article explores the intersection between statistics and artificial intelligence, focusing o... more This article explores the intersection between statistics and artificial intelligence, focusing on time series analysis and the prediction of extreme events such as climate catastrophes, material failures, and other rare occurrences. Time series analysis is widely used to forecast future behaviours based on historical data, but it faces significant challenges in predicting extremes due to its reliance on linear models that do not adequately capture the complex dynamics of such events. The limitations of linear and logistic regression models in forecasting maxima and minima are discussed, and nonlinear alternatives, including neural networks and extreme value models, are presented as offering greater accuracy and robustness in predicting rare events. The article concludes that, despite the difficulties, integrating advanced statistical methods with machine learning techniques can substantially improve the ability to forecast extreme events, contributing to better risk management and decision-making in critical scenarios.
This article explores the intersection between statistics and artificial intelligence, focusing o... more This article explores the intersection between statistics and artificial intelligence, focusing on time series analysis and the prediction of extreme events such as climate catastrophes, material failures, and other rare occurrences. Time series analysis is widely used to forecast future behaviours based on historical data, but it faces significant challenges in predicting extremes due to its reliance on linear models that do not adequately capture the complex dynamics of such events. The limitations of linear and logistic regression models in forecasting maxima and minima are discussed, and nonlinear alternatives, including neural networks and extreme value models, are presented as offering greater accuracy and robustness in predicting rare events. The article concludes that, despite the difficulties, integrating advanced statistical methods with machine learning techniques can substantially improve the ability to forecast extreme events, contributing to better risk management and decision-making in critical scenarios.
This article explores the intersection between statistics and artificial intelligence, focusing o... more This article explores the intersection between statistics and artificial intelligence, focusing on time series analysis and the prediction of extreme events such as climate catastrophes, material failures, and other rare occurrences. Time series analysis is widely used to forecast future behaviours based on historical data, but it faces significant challenges in predicting extremes due to its reliance on linear models that do not adequately capture the complex dynamics of such events. The limitations of linear and logistic regression models in forecasting maxima and minima are discussed, and nonlinear alternatives, including neural networks and extreme value models, are presented as offering greater accuracy and robustness in predicting rare events. The article concludes that, despite the difficulties, integrating advanced statistical methods with machine learning techniques can substantially improve the ability to forecast extreme events, contributing to better risk management and decision-making in critical scenarios.
This article explores the intersection between statistics and artificial intelligence, focusing o... more This article explores the intersection between statistics and artificial intelligence, focusing on time series analysis and the prediction of extreme events such as climate catastrophes, material failures, and other rare occurrences. Time series analysis is widely used to forecast future behaviours based on historical data, but it faces significant challenges in predicting extremes due to its reliance on linear models that do not adequately capture the complex dynamics of such events. The limitations of linear and logistic regression models in forecasting maxima and minima are discussed, and nonlinear alternatives, including neural networks and extreme value models, are presented as offering greater accuracy and robustness in predicting rare events. The article concludes that, despite the difficulties, integrating advanced statistical methods with machine learning techniques can substantially improve the ability to forecast extreme events, contributing to better risk management and decision-making in critical scenarios.