Estimating the Transitional Probabilities of the E.Coli Gene Chain by Maximum Likelihood Method and Bayes Method
Abstract
The transition matrix estimators of the Markov chain are not accurate and the transition matrix is considered given. There are many methods that are used to estimate the transition probabilities matrix for different cases, the most famous of which is the Maximum Likelihood Method, In order to find a good estimator for the transition probabilities matrix of the Markov chain, a Bayes method and a Proposed Method was used in this paper, to reach the transition probabilities with the least variance, The Escherichia Coli (E.Coli) gene chain was chosen as an applied aspect of the study due to its importance in medical research and for the purpose of discovering and manufacturing treatments by knowing the final form of its gene chain. After testing the E.Coli gene chain, it was found that is represents a Markov chain, and then both the transition probabilities matrix and the transition probabilities variance were estimated used Proposed Method and Bayes method and Maximum Likelihood Method, and it was found that the Proposed Method for transitional probabilities is better than the Bayes method and Maximum Likelihood Method dependence on the variance.
References
- Demirel, A., & elik, H. (2018). DIRICHLET DISTRIBUTION AND ESTIMATION OF PARAMETERS. ADVANCES APPLICATIONS IN STATISTICS, 53(4).
- Fister, I., Fister Jr, I., Yang, X.-S., & Brest, J. (2013). A comprehensive review of firefly algorithms. %J Swarm Evolutionary Computation 13, 34-46.
- Guo, H. R., Liao, H., Zhao, W., & Mettas, A. J. I. T. o. R. (2007). A new stochastic model for systems under general repairs. 56(1), 40-49.
- Kalbfleisch, J. D., & Lawless, J. F. (1984). Leastsquares estimation of transition probabilities from aggregate data. Canadian Journal of Statistics, 12(3), 169-182.
- Lee, T. C., Judge, G., & Zellner, A. (1968). Maximum likelihood and Bayesian estimation of transition probabilities. Journal of the American Statistical Association, 63(324), 1162-1179.
- Lin, J. (2016). On the dirichlet distribution. Department of Mathematics Statistics, Queens University 10-11.
- Ma, J., Yu, X., Symanski, E., Doody, R., & Chan, W. (2014). A Bayesian Approach in Estimating Transition Probabilities of a Discrete-time Markov Chain for Ignorable Intermittent Missing Data. Communications in Statistics-Simulation and Computation, 45(7), 2598-2616.
- Singer, P., Helic, D., Taraghi, B., & Strohmaier, M. J. P. o. (2014). Detecting memory and structure in human navigation patterns using Markov chain models of varying order. 9(7), e102070.
- Soyer, R., & Tarimcilar, M. M. (2008). Modeling and analysis of call center arrival data: A Bayesian approach. %J Management Science, 54(2), 266-278.
- Strelioff, C. C., Crutchfield, J. P., & Hbler, A. W. J. P. R. E. (2007). Inferring Markov chains: Bayesian estimation, model comparison, entropy rate, and out-of-class modeling. 76(1), 011106.
- Wang, G. (2010). ML estimation of transition probabilities in jump Markov systems via convex optimization. IEEE Transactions on Aerospace Electronic Systems 46(3), 1492-1502.
- Yang, J., & Sha, Q. (2011). Research and application by Markov chain operators in the mobile phone market. Paper presented at the 2011 2nd International Conference on Artificial Intelligence, Management Science and Electronic Commerce (AIMSEC).
