Hidden Markov Models for Time Series
Hidden Markov Models for Time Series

Author: Walter Zucchini

Publisher: CRC Press

Published: 2017-12-19

Total Pages: 370

ISBN-13: 1482253844

DOWNLOAD EBOOK

Hidden Markov Models for Time Series: An Introduction Using R, Second Edition illustrates the great flexibility of hidden Markov models (HMMs) as general-purpose models for time series data. The book provides a broad understanding of the models and their uses. After presenting the basic model formulation, the book covers estimation, forecasting, decoding, prediction, model selection, and Bayesian inference for HMMs. Through examples and applications, the authors describe how to extend and generalize the basic model so that it can be applied in a rich variety of situations. The book demonstrates how HMMs can be applied to a wide range of types of time series: continuous-valued, circular, multivariate, binary, bounded and unbounded counts, and categorical observations. It also discusses how to employ the freely available computing environment R to carry out the computations. Features Presents an accessible overview of HMMs Explores a variety of applications in ecology, finance, epidemiology, climatology, and sociology Includes numerous theoretical and programming exercises Provides most of the analysed data sets online New to the second edition A total of five chapters on extensions, including HMMs for longitudinal data, hidden semi-Markov models and models with continuous-valued state process New case studies on animal movement, rainfall occurrence and capture-recapture data

Improving Time Structure Patterns of Orthogonal Markov Chains and Its Consequences in Hydraulic Simulations
Improving Time Structure Patterns of Orthogonal Markov Chains and Its Consequences in Hydraulic Simulations

Author: Juan Carlos Jaimes Correa

Publisher:

Published: 2013

Total Pages: 97

ISBN-13:

DOWNLOAD EBOOK

Rainfall (liquid precipitation) occurrences understood as rain events are relevant for agricultural practices because temporal distribution of rainfall highly affects yield production. A few stochastic models satisfactorily generate daily rainfall events while preserving temporal and spatial dependence among multiple sites. I evaluated an extension on the traditional Orthogonal Markov chain (TOMC) model in reproducing the temporal structure of rainfall events at multiple sites in Florida (FL), Nebraska (NE) and California (CA). In addition, a simulation of watershed runoff from rainfall events, reproduced by a single- and multi-site weather generator, was conducted. Results shows that (i) a temporal structure extended Orthogonal Markov chain (EOMC) maintained the spatial correlation between observed and generated rainfall events; (ii) EOMC used a smaller number of yearlong simulations for generating the observed frequencies of wet spells than TOMC requires for similar accuracy; (iii) using EOMC generated rainfall data in SWMM produced similar median runoff values to those generated using observed data; and (iv) EOMC reduces 50% of computing time for generating rainfall data. EOMC can benefit modeling of future climate scenarios by economical reduction of hardware need.

Crop Climate Simulation Modelling
Crop Climate Simulation Modelling

Author: M. Sayedur Rahman

Publisher: LAP Lambert Academic Publishing

Published: 2012-04

Total Pages: 220

ISBN-13: 9783848496228

DOWNLOAD EBOOK

A stochastic model based on first-order Markov chain was developed to simulate daily rainfall. The model is capable of simulating daily rainfall data of any length, based on the estimated transitional probabilities, mean, standard deviation and skew coefficients of rainfall amounts. A study of rainfall probability is an approach to sound planning for any variation of rainfall either small or large. The simulation model has been used successfully to estimate daily rainfall. The Multivariate logistic regression is used to estimate the probability that it is raining. The logistic regression technique is used to compare between the actual and simulation results for a rainfall from January to December in Bangladesh. The probability of occurrence of rainfall is of vital importance in efficient planning and execution of water use program. This study describes a crop-climate simulation model under rainfed conditions in Bangladesh to be used as a tool for analyzing growth and yield to help planning and management of rice production.

Stochastic Modeling of Daily Precipitation Process in the Context of Climate Change
Stochastic Modeling of Daily Precipitation Process in the Context of Climate Change

Author: Sarah El Outayek

Publisher:

Published: 2021

Total Pages:

ISBN-13:

DOWNLOAD EBOOK

"Information on the variations of rainfall in space and time is essential for the design and management of different water resources systems. This thesis proposed a new stochastic model (referred herein as the MCME model) that is able to capture accurately the statistical properties of the observed daily precipitation process for the current and future climates under different climate change scenarios. The MCME model consists of two components: (i) the first component representing the daily precipitation occurrence process based on the first-order two-state Markov Chain (MC); and (ii) the second component describing the distribution of daily precipitation amounts using the Mixed Exponential (ME) distribution. A comparative study was carried out to assess the performance of the proposed model as compared to the popular LARS-WG model using observed daily precipitation data from a network of nine raingauges representing different climatic conditions across Quebec. Results of this study have indicated the better performance of the MCME model in terms of its accuracy and robustness in the modeling of the daily precipitation process. In addition, an improved perturbation method was developed for establishing the linkages between the proposed MCME model with the coarse-scale climate model outputs. Results of a comparative study using both MCME and LARS-WG models have demonstrated the best performance of the proposed perturbation method as compared with other existing perturbation methods in terms of its accuracy in capturing different statistical properties of the projected daily precipitation process for future periods. Finally, an assessment of the performance of the MCME and LARS-WG models based on the proposed perturbation technique was performed in the context of climate change using daily precipitation data from a network of five stations located in Quebec and Ontario and the downscaled simulation data from 21 different global climate models. Results of this assessment have indicated the feasibility and accuracy of the proposed MCME model and the proposed perturbation technique for downscaling daily precipitation processes for impact and adaptation studies in practice"--