Lifetime Markov Model Dynamic Cohorts
Tan Nguyen
Hi everyone,
For those interested in building dynamic cohort Markov models with probabilistic analysis, I noticed TreeAge cannot look-up the different age groups using distribution functions. Instead, you will need to create each age cohort Markov model individually and input the parameters individually. It is time consuming but works a treat. Happy to help anyone looking to do this.
Comments
As someone who supports TreeAge Pro, I am interested in what you did.
Hi Andrew,
Essentially, for each age-cohort, I created the appropriate Markov model. If you decide to use a probabilistic analysis where each age cohort probabilities and payoffs are different, you will need to create each parameter using the distribution function. Table functions work well with the distribution functions. Input all parameters in the tree, and then once you are happy with it, use clones to copy Markov model. This serves as a template, which you can then add more age-cohorts. The clones can then be destroyed so that you can modify each parameter which is relevant to each age cohort.
Cheers.
Thank you.
Please sign in to leave a comment.