Markov Models

Markov Models

1. Semi Markov models

   • Drops assumptions of transition times no longer exponentially
   • More realistic
   • extremely difficult to solve

2. AD/DIS

   • ADVANTAGES
     • solved exactly
     • straightfoward
     • use clinical staging systems available from studies
     • Model the maths and diagrams easily
     • easy to understand
     • several outputs
   • DISADVANTAGES
     • memoryless
     • life isn't exponential
     • difficult to choose time scales
     • discrete
     • too many assumptions

3. EXPLAIN

   • state transition diagram
   • absorbing, recurrent, transient

4. FIND

   • limiting behaviour
   • time to certain states

5. NEED

   • time- discrete or continuous
   • transition probabilities
   • memoryless
   • estimate the parameters
   • exponential dwelling times