Search results

  1. staassis

    Good book on markov chains

    Chapters 4-7 of Ross, S. M. (2009). Introduction to Probability Models (10th ed). Academic Press. will serve as a rigorous introduction into Markov chains and closely related stochastic processes. Later on, when you solve actual problems, you may fall back on Lawler, G. F. (1995)...
  2. staassis

    Good book on markov chains

    Noetsi, so are you interested in 1) an introductory book where all major properties of discrete-time and continuous-time Markov chains are nicely laid out or 2) a more advanced book showing applications of Markov chains to real-life processes? Or are you interested in most general Markov...
  3. staassis

    Markov chain with covariates

    That is why I mentioned: "You would just use the idea for the algorithm, not the functional form." Just apply the algorithm to various defining characteristics of the Markov chain.
  4. staassis

    What is the formula for estimating the standard error of variance ratio?

    The sample variance ratio divided by the true variance ratio will have F-distribution if the underlying data are Gaussian. Use that.
  5. staassis

    Markov chain with covariates

    This paper by Breiman gives you an idea of how to sample just a subset of features and observations at each step and merge the result into the existing model. You would just use the idea for the algorithm, not the functional form, of course...
  6. staassis

    Markov chain with covariates

    A word of advice: you can try making this shorter for more people to read and respond. Currently this looks like a director's cut of the Lord of the Rings... Having that said, if msm cannot handle your data, you will have to do programming of your own, with some parallel computing perhaps...
  7. staassis

    How to compare two sets of proportions?

    Take a look at the Stata document that I have linked... The data do not seem to be paired, based on your description.
  8. staassis

    How to compare two sets of proportions?

    Why "paired"? Also a t-test would not use the information on the reported standard errors. You should use methods of meta-analysis, fixed effects and random effects models in particular.
  9. staassis

    Ranking stocks based on EPS growth rates

    Not quite. It is better to consider t-statistic = [Average EPS] / [Standard Error of Average EPS]. Such ranking would be related to the popular procedure where assets are ranked using the Sharpe ratios associated with trading them.
  10. staassis


    Correct... as long as you have calculated the sums correctly.
  11. staassis


    For each subskill i you need to create variable X_i = Measurement After - Measurement Before. All analysis should be run on these differences only. If the sample size is substantial, it is best to run 6 univariate ANOVA and apply the Bonferroni correction. If the sample size is modest (like...
  12. staassis

    Best way to test for a relationship between rainfall and river discharge?

    A time series model with lags and/or seasonalities.
  13. staassis

    How to model this problem?

    You seem to imply that there is a reflexive relationship between Y on the one hand and {X1, X2} on the other hand. If that is the case, one framework for studying the mechanism of impacts is Structural Vector Autoregression (SVAR). It can be rephrased in an equivalent VAR form.
  14. staassis

    Survival Analysis

    The effects of predictors on the survival curves is accounted for in the aforementioned models.
  15. staassis

    Survival Analysis

    Since there are confounders to adjust for, you need to use semi-parametric methods of survival analysis, like Cox proportional hazards model, accelerated failure time model, etc. Whether each parameter equals 0 can be determined using the corresponding Wald test or likelihood ratio test,
  16. staassis

    Time Series books

    Noetsi, you can also try (in case you have not done that already). A good place to ask questions about statistical literature.
  17. staassis

    Time Series books

    I think this is an interesting topic. I did put my 2 cents in but would anybody else have anything to contribute?
  18. staassis

    Probability of regression result

    If it's not an idealized hw problem, then linear regression is far from being appropriate. First, you need to transform the data into returns, log-returns or some other weakly stationary stochastic process. Second, you need to apply time series techniques. In particular, as ondansetron pointed...
  19. staassis

    To assess meaningfulness between an independent variable and a dependent one, is just the p value alone adequate and/or the correlation?

    The two characterizations are separate: 1] "The effect is statistically significant" means that the true parameter has a value different from the value in the null hypothesis. In the modeling setting, this typically implies that the corresponding predictor has a relationship with the dependent...
  20. staassis

    Which statistical test to use

    "I feel it is sufficient to answer the question..." You feel? I thought your coming to this forum was a demonstration of understanding that you do not know statistics sufficiently well...