Examinando por Materia "ANALISIS DE SERIES DE TIEMPO"
Mostrando1 - 15 de 15
Resultados por página
Opciones de clasificación
- Trabajo final de especializaciónAnálisis de series de tiempo: pronóstico de demanda de uso de aeropuertos en Argentina al 2022(2018) López Sáez, José Ignacio; Gambini, Juliana"En la mayoría de los negocios, se desea ser capaz de estimar la demanda futura de un producto o servicio dado. El análisis sobre series temporales permite utilizar la información histórica para ofrecer un número aproximado de dicho valor, dentro de un rango de probabilidades determinado. Este estudio surge a partir de la necesidad que tiene el Ministerio de Transporte de la Nación de conocer el número de pasajeros que utilizarán cada aeropuerto del país en el futuro para poder asignar de una manera más eficiente los recursos disponibles y orientar inversiones. En este estudio se han evaluado en total 4 modelos de proyección de series de tiempo: suavizamiento exponencial (Holt-Winters), ARIMA, Prophet (desarrollado por Facebook) y redes neuronales (procedimientos de aprendizaje automático o machine learning), para las cuales se probaron tres implementaciones distintas en R. Se obtiene así el resultado de proyección de pasajeros domésticos e internacionales para cada una de estas seis implementaciones y para todos los aeropuertos de la Argentina para un horizonte de 5 años (60 meses a partir del último disponible), entregando también el error de ajuste de cada uno de los modelos."
- Artículo de Publicación PeriódicaBandt-Pompe symbolization dynamics for time series with tied values: a data-driven approach(2018-07) Traversaro Varela, Francisco; Redelico, Francisco; Risk, Marcelo; Frery, Alejandro C.; Rosso, Osvaldo A."In 2002, Bandt and Pompe [Phys. Rev. Lett. 88, 174102 (2002)] introduced a successfully symbolic encoding scheme based on the ordinal relation between the amplitude of neighboring values of a given data sequence, from which the permutation entropy can be evaluated. Equalities in the analyzed sequence, for example, repeated equal values, deserve special attention and treatment as was shown recently by Zunino and co-workers [Phys. Lett. A 381, 1883 (2017)]. A significant number of equal values can give rise to false conclusions regarding the underlying temporal structures in practical contexts. In the present contribution, we review the different existing methodologies for treating time series with tied values by classifying them according to their different strategies. In addition, a novel data-driven imputation is presented that proves to outperform the existing methodologies and avoid the false conclusions pointed by Zunino and co-workers."
- Artículo de Publicación PeriódicaCharacterization of autoregressive processes using entropic quantifiers(2018-01) Traversaro Varela, Francisco; Redelico, Francisco"The aim of the contribution is to introduce a novel information plane, the causal-amplitude informational plane. As previous works seems to indicate, Bandt and Pompe methodology for estimating entropy does not allow to distinguish between probability distributions which could be fundamental for simulation or for probability analysis purposes. Once a time series is identified as stochastic by the causal complexity-entropy informational plane, the novel causal-amplitude gives a deeper understanding of the time series, quantifying both, the autocorrelation strength and the probability distribution of the data extracted from the generating processes. Two examples are presented, one from climate change model and the other from financial markets"
- Artículo de Publicación PeriódicaConfidence intervals and hypothesis testing for the Permutation Entropy with an application to epilepsy(2018-04) Traversaro Varela, Francisco; Redelico, Francisco"In nonlinear dynamics, and to a lesser extent in other fields, a widely used measure of complexity is the Permutation Entropy. But there is still no known method to determine the accuracy of this measure. There has been little research on the statistical properties of this quantity that characterize time series. The literature describes some resampling methods of quantities used in nonlinear dynamics - as the largest Lyapunov exponent - but these seems to fail. In this contribution, we propose a parametric bootstrap methodology using a symbolic representation of the time series to obtain the distribution of the Permutation Entropy estimator. We perform several time series simulations given by well-known stochastic processes: the 1/f α noise family, and show in each case that the proposed accuracy measure is as efficient as the one obtained by the frequentist approach of repeating the experiment. The complexity of brain electrical activity, measured by the Permutation Entropy, has been extensively used in epilepsy research for detection in dynamical changes in electroencephalogram (EEG) signal with no consideration of the variability of this complexity measure. An application of the parametric bootstrap methodology is used to compare normal and pre-ictal EEG signals."
- Artículo de Publicación PeriódicaCrude oil market and geopolitical events: an analysis based on information-theory-based quantifiers(2016) Fernández Bariviera, Aurelio; Zunino, Luciano; Rosso, Osvaldo A."This paper analyzes the informational efficiency of oil market during the last three decades, and examines changes in informational efficiency with major geopolitical events, such as terrorist attacks, financial crisis and other important events. The series under study is the daily prices of West Texas Intermediate (WTI) in USD/BBL, commonly used as a benchmark in oil pricing. The analysis is performed using information-theory-derived quantifiers, namely permutation entropy and permutation statistical complexity. These metrics allow capturing the hidden structure in the market dynamics, and allow discriminating different degrees of informational efficiency. We find that some geopolitical events impact on the underlying dynamical structure of the market."
- Proyecto final de GradoCryptobot: creación de base de datos y estudio de series temporales de criptomonedas(2019-12-18) Oviedo Candelaresi, Fernán Darío; Fratoni, Axel; Parisi, Daniel"Existe una plataforma de intercambio de criptomonedas llamada HitBTC que provee una API para interactuar con los mercados. El objetivo de este trabajo fue desarrollar un sistema de software que permitiera recabar información del mercado de forma automática a lo largo del tiempo, analizar esta información para caracterizar el comportamiento del mercado e implementar estrategias de decisión automática para ejecutar operaciones de compraventa."
- Trabajo final de especializaciónDetección de puntos de cambio en la tendencia de la fecundidad en Argentina entre 2011 y 2017(2020-09-01) Guevel, Carlos Gustavo; Gambini, Juliana"Este trabajo tiene su origen en un estudio preliminar descriptivo realizado en el año 2019 en el ámbito de la Secretaría de Gobierno de Salud, a raíz de una inquietud planteada por el Programa de Salud Sexual y Reproductiva, que contó con apoyo de UNFPA y el Instituto de Salud Colectiva de la Universidad Nacional de Lanús (informe no publicado). El mismo fue motivado por el estudio realizado en Uruguay. Quedó pendiente en esa primera aproximación el análisis cuantitativo de los cambios y solo se consideró el grupo de edad 15 a 24 años. El presente estudio busca proporcionar evidencia a partir del análisis cuantitativo para aportar a la discusión sobre la reducción de la fecundidad ocurrida en los últimos años en Argentina y que los hallazgos sirvan para la orientación de las políticas e intervenciones sobre salud sexual y reproductiva."
- Trabajo final de especializaciónEstimación de tiempos de espera en peajes(2019-08-16) Ailán, Julián; Gambini, Juliana"En el presente trabajo práctico final se aborda el estudio del estado del tránsito en autopistas concesionadas a Autopistas Urbanas Sociedad Anónima (AUSA),en la Ciudad Autónoma de Buenos Aires haciendo énfasis en el caudal vehicular que transita a través de los peajes ubicados en estas autopistas."
- Artículo de Publicación PeriódicaLibor at crossroads: stochastic switching detection using information theory quantifiers(2016-07) Fernández Bariviera, Aurelio; Guercio, M. Belén; Martinez, Lisana B.; Rosso, Osvaldo A."This paper studies the 28 time series of Libor rates, classified in seven maturities and four currencies, during the last 14 years. The analysis was performed using a novel technique in financial economics: the Complexity-Entropy Causality Plane. This planar representation allows the discrimination of different stochastic and chaotic regimes. Using a temporal analysis based on moving windows, this paper unveils an abnormal movement of Libor time series around the period of the 2007 financial crisis. This alteration in the stochastic dynamics of Libor is contemporary of what press called "Libor scandal", i.e. the manipulation of interest rates carried out by several prime banks. We argue that our methodology is suitable as a market watch mechanism, as it makes visible the temporal redution in informational efficiency of the market."
- Ponencia en CongresoPatterns in temporal series of meteorological variables using SOM & TDIDT(2006) Cogliati, Marisa; Britos, Paola Verónica; García Martínez, Ramón"The purpose of the present article is to investigate if there exist any such set of temporal stable patterns in temporal series of meteorological variables studying series of air temperature, wind speed and direction an atmospheric pressure in a period with meteorological conditions involving nocturnal inversion of air temperature in Allen, Rio Negro, Argentina. Our conjecture is that there exist independent stable temporal activities, the mixture of which give rise to the weather variables; and these stable activities could be extracted by Self Organized Maps plus Top Down Induction Decision Trees analysis of the data arising from the weather patterns, viewing them as temporal signals."
- Artículo de Publicación PeriódicaPermutation entropy based time series analysis: equalities in the input signal can lead to false conclusions(2017-06) Zunino, Luciano; Olivares, Felipe; Scholkmann, Felix; Rosso, Osvaldo A."A symbolic encoding scheme, based on the ordinal relation between the amplitude of neighboring values of a given data sequence, should be implemented before estimating the permutation entropy. Consequently, equalities in the analyzed signal, i.e. repeated equal values, deserve special attention and treatment. In this work, we carefully study the effect that the presence of equalities has on permutation entropy estimated values when these ties are symbolized, as it is commonly done, according to their order of appearance. On the one hand, the analysis of computer-generated time series is initially developed to understand the incidence of repeated values on permutation entropy estimations in controlled scenarios. The presence of temporal correlations is erroneously concluded when true pseudorandom time series with low amplitude resolutions are considered. On the other hand, the analysis of real-world data is included to illustrate how the presence of a significant number of equal values can give rise to false conclusions regarding the underlying temporal structures in practical contexts."
- Artículo de Publicación PeriódicaPermutation min-entropy: an improved quantifier for unveiling subtle temporal correlations(2015-01) Zunino, Luciano; Olivares, Felipe; Rosso, Osvaldo A."The aim of this letter is to introduce the permutation min-entropy as an improved symbolic tool for identifying the existence of hidden temporal correlations in time series. On the one hand, analytical results obtained for the fractional Brownian motion stochastic model theoretically support this hypothesis. On the other hand, the analysis of several computer-generated and experimentally observed time series illustrate that the proposed symbolic quantifier is a versatile and practical tool for identifying the presence of subtle temporal structures in complex dynamical systems. Comparisons against the results obtained with other tools confirm its usefulness as an alternative and/or complementary measure of temporal correlations."
- Artículo de Publicación PeriódicaA simple and fast representation space for classifying complex time series(2017-03) Zunino, Luciano; Olivares, Felipe; Fernández Bariviera, Aurelio; Rosso, Osvaldo A."In the context of time series analysis considerable effort has been directed towards the implementation of efficient discriminating statistical quantifiers. Very recently, a simple and fast representation space has been introduced, namely the number of turning points versus the Abbe value. It is able to separate time series from stationary and non-stationary processes with long-range dependences. In this work we show that this bidimensional approach is useful for distinguishing complex time series: different sets of financial and physiological data are efficiently discriminated. Additionally, a multiscale generalization that takes into account the multiple time scales often involved in complex systems has been also proposed. This multiscale analysis is essential to reach a higher discriminative power between physiological time series in health and disease. "
- Proyecto final de GradoSistema automático para compraventa de activos financieros(2018) Mounier, Agustín; Suárez Bodean, Joaquín; Parisi, Daniel"En este trabajo nos centraremos en el comercio de acciones, uno de los valores más conocidos y comunes con los que opera el mercado. Probablemente esté familiarizado con la definición popular de acción como: “Una acción es una parte de la empresa y al ser de su propiedad tiene derecho a reclamar parte de las ganancias de la misma.” Desafortunadamente, esta definición es incorrecta en algunos aspectos."
- Trabajo final de especializaciónUtilización de redes neuronales recurrentes en la predicción de tendencias del mercado de harina de soja(2022) Alonso, Juan Ignacio; Riccillo, Marcela"La volatilidad de los precios internacionales de la Harina de Soja impacta de manera significativa la economía de distintas industrias, gobiernos y, finalmente, la población. Los modelos Auto Regresivos de Media Móvil (ARIMA) constituyen una de las herramientas de análisis de series de tiempo más utilizadas. Sin embargo, el advenimiento de nuevas tecnologías de análisis y procesamiento de datos difundieron nuevas técnicas aplicables al estudio de series de tiempo, siendo Las Redes Neuronales Recurrentes del tipo LSTM una de ellas. En el presente estudio se compara la performance relativa de modelos ARIMA y RNR LSTM en la predicción de tendencias de precios de Harina de Soja."