Daily Dozen:
1. What is the main goal of the Baum-Welch algorithm?
2. Why can't we find the optimal solution for finding the parameters that maximize P(x|params, model)?
3. What does "locally optimal" mean?
4. Do we have to assume a particular HMM topology to use Baum-Welch?
5. What sort of prior knowledge might we want to use to choose the initial parameters for Baum-Welch?
6. Why were the first gene finders for prokaryotes (other than their simplicity)?
7. How many characters/states does an nth-order Markov model consider?
8. What is an Interpolated Markov Model?
9. Why is eukaryotic gene finding so difficult?
10. What is Maximal Dependence Decomposition?
11. What advantages would a dual- or multiple-genome predictor confer?
12. What is the current rate of correct exon prediction in eukaryotes? Prokaryotes?
1. What is the main goal of the Baum-Welch algorithm?
2. Why can't we find the optimal solution for finding the parameters that maximize P(x|params, model)?
3. What does "locally optimal" mean?
4. Do we have to assume a particular HMM topology to use Baum-Welch?
5. What sort of prior knowledge might we want to use to choose the initial parameters for Baum-Welch?
6. Why were the first gene finders for prokaryotes (other than their simplicity)?
7. How many characters/states does an nth-order Markov model consider?
8. What is an Interpolated Markov Model?
9. Why is eukaryotic gene finding so difficult?
10. What is Maximal Dependence Decomposition?
11. What advantages would a dual- or multiple-genome predictor confer?
12. What is the current rate of correct exon prediction in eukaryotes? Prokaryotes?