
Ce travail introduit une méthodologie pour estimer les prix du café en fonction de l'utilisation de machines d'apprentissage extrême. Le processus est initié en identifiant la présence de composantes non stationnaires, telles que la saisonnalité et la tendance. Ces composants sont retirés s'ils sont trouvés. Ensuite, les décalages temporels sont sélectionnés en fonction de la réponse du filtre de la fonction d'autocorrélation partielle. En tant que prédicteurs, nous abordons les modèles suivants : les modèles Exponential Smoothing (ES), Autoregressive (AR) et Autoregressive Integrated and Moving Average (ARIMA), les réseaux neuronaux Multilayer Perceptron (MLP) et Extreme Learning Machines (ELM). Les résultats de calcul basés sur trois mesures d'erreur et deux types de café (Arabica et Robusta) ont montré que les réseaux de neurones, en particulier l'ORME, peuvent atteindre des niveaux de performance plus élevés que les autres modèles. La méthodologie, qui présente les étapes de prétraitement, la sélection du retard et l'utilisation de L'ORME, est une nouveauté qui contribue au domaine de la prévision des prix du café.
Este trabajo introduce una metodología para estimar los precios del café basada en el uso de Máquinas Extremas de Aprendizaje. El proceso se inicia identificando la presencia de componentes no estacionarios, como la estacionalidad y la tendencia. Estos componentes se retiran si se encuentran. A continuación, se seleccionan los retardos temporales en función de la respuesta del filtro Función de autocorrelación parcial. Como predictores, abordamos los siguientes modelos: modelos de Suavizado Exponencial (ES), Autoregresivo (AR) y Autoregresivo Integrado y Media Móvil (ARIMA), Perceptrón Multicapa (MLP) y redes neuronales de Máquinas de Aprendizaje Extremo (ELM). Los resultados computacionales basados en tres métricas de error y dos tipos de café (Arábica y Robusta) mostraron que las redes neuronales, especialmente el OLMO, pueden alcanzar niveles de rendimiento más altos que los otros modelos. La metodología, que presenta etapas de preprocesamiento, selección de lag y uso de ELM, es una novedad que contribuye al campo de la previsión de precios del café.
This work introduces a methodology to estimate coffee prices based on the use of Extreme Learning Machines. The process is initiated by identifying the presence of nonstationary components, like seasonality and trend. These components are withdrawn if they are found. Next, the temporal lags are selected based on the response of the Partial Autocorrelation Function filter. As predictors, we address the following models: Exponential Smoothing (ES), Autoregressive (AR) and Autoregressive Integrated and Moving Average (ARIMA) models, Multilayer Perceptron (MLP) and Extreme Learning Machines (ELMs) neural networks. The computational results based on three error metrics and two coffee types (Arabica and Robusta) showed that the neural networks, especially the ELM, can reach higher performance levels than the other models. The methodology, which presents preprocessing stages, lag selection, and use of ELM, is a novelty that contributes to the coffee prices forecasting field.
يقدم هذا العمل منهجية لتقدير أسعار القهوة بناءً على استخدام آلات التعلم المتطرف. تبدأ العملية من خلال تحديد وجود مكونات غير ثابتة، مثل الموسمية والاتجاه. يتم سحب هذه المكونات إذا تم العثور عليها. بعد ذلك، يتم تحديد التأخيرات الزمنية بناءً على استجابة مرشح وظيفة الارتباط الذاتي الجزئي. كمتنبئين، نتناول النماذج التالية: التنعيم الأسي (ES)، والنماذج الانحدارية الذاتية (AR)، ونماذج المتوسطات المتحركة والمتكاملة الانحدارية (ARIMA)، والشبكات العصبية المدركة متعددة الطبقات (MLP) وآلات التعلم المتطرفة (ELMs). أظهرت النتائج الحسابية القائمة على ثلاثة مقاييس للخطأ ونوعين من القهوة (أرابيكا وروبوستا) أن الشبكات العصبية، وخاصة علم، يمكن أن تصل إلى مستويات أداء أعلى من النماذج الأخرى. المنهجية، التي تقدم مراحل ما قبل المعالجة، واختيار التأخر، واستخدام علم، هي حداثة تساهم في مجال التنبؤ بأسعار القهوة.
Artificial intelligence, Coffee price forecasting, Extreme learning machine, Agriculture (General), Social Sciences, Autoregressive model, S1-972, Decision Sciences, Engineering, Theory and Applications of Extreme Learning Machines, Ensemble Learning, Artificial neural networks, Extreme value theory, Statistics, T58.5-58.64, Exponential smoothing, Autocorrelation, Physical Sciences, Artificial neural network, Time series, Electricity Price and Load Forecasting Methods, Information technology, Management Science and Operations Research, Forecasting Models, FOS: Economics and business, Artificial Intelligence, Moving average, Machine learning, FOS: Electrical engineering, electronic engineering, information engineering, FOS: Mathematics, Multilayer perceptron, Econometrics, Electrical and Electronic Engineering, Computational intelligence, Autoregressive integrated moving average, Electricity Price Forecasting, Predicting Stock Market Trends and Movements, Computer science, Partial autocorrelation function, Computer Science, Computer vision, Linear models, Extreme Learning Machine, Short-Term Forecasting, Mathematics
Artificial intelligence, Coffee price forecasting, Extreme learning machine, Agriculture (General), Social Sciences, Autoregressive model, S1-972, Decision Sciences, Engineering, Theory and Applications of Extreme Learning Machines, Ensemble Learning, Artificial neural networks, Extreme value theory, Statistics, T58.5-58.64, Exponential smoothing, Autocorrelation, Physical Sciences, Artificial neural network, Time series, Electricity Price and Load Forecasting Methods, Information technology, Management Science and Operations Research, Forecasting Models, FOS: Economics and business, Artificial Intelligence, Moving average, Machine learning, FOS: Electrical engineering, electronic engineering, information engineering, FOS: Mathematics, Multilayer perceptron, Econometrics, Electrical and Electronic Engineering, Computational intelligence, Autoregressive integrated moving average, Electricity Price Forecasting, Predicting Stock Market Trends and Movements, Computer science, Partial autocorrelation function, Computer Science, Computer vision, Linear models, Extreme Learning Machine, Short-Term Forecasting, Mathematics
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 28 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Top 10% | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Top 10% |
