654 reputation
410
bio website
location
age
visits member for 3 years, 10 months
seen yesterday

I'm an ABD statistician, and work fulltime in this capacity. I have interest in Computational Statistics, Markov Chain Theory, and Sample Survey.


Oct
14
revised Understanding $O_p$
spelling
Oct
14
revised Understanding $O_p$
added 124 characters in body
Oct
14
revised Understanding $O_p$
added T distribution
Oct
14
answered Understanding $O_p$
Oct
13
comment Performance drop between training and validation datasets
Could you re-partition the data sets and try again? What are the sizes of your data sets? For relatively small sample sizes this may just be sampling error. You might also want to define want mean error means, I assume you mean Mean Absolute Error or Mean Squared Error? Personally I would take a look at some resampling such as k-fold validation to obtain a slightly more robust estimate.
Oct
8
awarded  Yearling
Oct
8
comment If $\frac{(2n-1)s^2}{\sigma^2} \overset{}{\sim} \chi^2_{2n-1}$, can I take the variance of both sides to get an equality relation?
This is fine if the sample is normally distributed, and the result will hold under the conditions that the chi-squared distribution holds. Note that you can't take the variance of $~\chi^2_{2n-1}$ because it isn't a random variable it's a statement of a distribution.
Oct
8
answered Jointly sufficient statistic?
Oct
8
answered Is kNN best for classification?
Oct
7
awarded  Explainer
Oct
7
comment Poisson as a limiting case of negative binomial
Consider the simple case of factorials. $\frac{ \prod_{i=1}^k (r+i-1)...(r-1) } { \prod_{i=1}^k (r+\lambda) }$ for bounded $k$ and $\lambda$ this goes to 1 as $r$ goes to $\infty$.
Oct
7
revised Poisson as a limiting case of negative binomial
spelling and grammer
Oct
7
suggested suggested edit on Poisson as a limiting case of negative binomial
Oct
7
answered Poisson as a limiting case of negative binomial
Oct
7
comment How to check the convergence of a binary simulation process?
convergence to what, and in what sense?
Oct
7
answered $Var(\bar{X})$ for a random sample from Bernoulli Distribution
Oct
6
answered How to choose the right model after k-fold cross validation is done?
Oct
6
comment How can I make sure that an LDA implementation works?
What output from the LDA are you dumping into your neural network? Also note that LDA is going to break the feature space by hyperplanes, if your groups are not easily separated by hyperplanes (consider a donut hole inside a donut) then LDA isn't going to help much. If you can, you might want to do some plots of your important features first, or some random features first to get a feel for your data.
Sep
29
comment are qqplots appropriate for time series?
This question is still pretty confusing. I'm not sure what you mean by the ordering of the QQ plot. The data in the QQ plot are ordered, which makes sense because they are quantiles. But the ordering should not be related to the index of the time series. However if you are worried about the stationarity then plot the ACF of the residuals. I don't understand the edit portion, the finite sum of normal R.V.s are normal, so if they aren't normal then your assumption was wrong.
Sep
29
comment smart sampling techniques in r
@user49422, what you are going to have to do is to determine what miss-classification rate or variance you are satisfied with. If this is simply as low as possible give your computing resources then you need to figure out what you can do given your computing resources. So try a few increasing sample sizes, plot out the compute time, memory use, and classification rate as a function of your sample size; then extrapolate from this what a reasonable 'max sample size' given your time/compute constraints would be.