Vyshnavi, M and Muthukumar, M (2025) HIDDEN MARKOV MODELS FOR ESTIMATING IMPORT DYNAMICS WITH PROBABILITY DISTRIBUTION ANALYSIS. HIDDEN MARKOV MODELS FOR ESTIMATING IMPORT DYNAMICS WITH PROBABILITY DISTRIBUTION ANALYSIS, 20 (2(84)).
![[thumbnail of HIDDEN MARKOV MODELS FOR ESTIMATING IMPORT DYNAMICS WITH PROBABILITY DISTRIBUTION ANALYSIS.pdf]](https://ir.psgcas.ac.in/style/images/fileicons/text.png)
Text
HIDDEN MARKOV MODELS FOR ESTIMATING IMPORT DYNAMICS WITH PROBABILITY DISTRIBUTION ANALYSIS.pdf - Published Version
Download (1MB)
Abstract
This article examines the use of Hidden Markov Models and probability distributions to study
agricultural import dynamics, with a focus on revealing hidden patterns in the data. The idea behind
the study is to enhance our understanding of how transitions between different agricultural import
states occur and to explore the most suitable probability distributions for modeling these hidden
states. The purpose of the study is to identify the optimal distributional fit for hidden states by
evaluating the Transition Probability Matrix, Emission Probability Matrix, and Initial Probability
Vector (π) for each state. The research implements the Akaike Information Criterion and the Bayesian
Information Criterion to select the best-fitting distribution for each scenario. Furthermore, the paper
focuses on the practical implications of these discoveries, such as determining the most likely state
sequence using the Viterbi path, which can help influence decision-making and forecasting. The
analysis is carried out using R software, which provides information about the associations between
probability distributions, stationarity tests, and the role of model selection criteria such as AIC and
BIC. Graphical representations of AIC and BIC values over several probability distributions, and
additionally a correlation matrix between selected distributions, help to highlight the findings.
Overall, the paper enhances our understanding of probability distributions through HMM
frameworks for agricultural import dynamics, providing recommendations for optimal model
selection in various applications.
| Item Type: | Article |
|---|---|
| Uncontrolled Keywords: | Probability Distributions, Hidden Markov Model, Lomax Distribution, Viterbi Path, Correlation Matrix. |
| Divisions: | PSG College of Arts and Science > Department of Statistics |
| Depositing User: | Dr. B Sivakumar |
| Date Deposited: | 31 Oct 2025 05:58 |
| Last Modified: | 31 Oct 2025 05:58 |
| URI: | https://ir.psgcas.ac.in/id/eprint/2484 |
Actions (login required)
- View Item