meta user
network profile
| bio | website | |
|---|---|---|
| location | Salt Lake City | |
| age | ||
| visits | member for | 1 year, 7 months |
| seen | 10 hours ago | |
| stats | profile views | 211 |
Praneeth Vepakomma
Mathematical and Applied Statistics, Rutgers University
|
May 19 |
revised |
Why are eigen and svd decompositions of a covariance matrix based on sparse data yielding different results? added 39 characters in body |
|
May 17 |
revised |
Gibbs sampler from conditional distribution added 1 characters in body |
|
May 12 |
awarded | Civic Duty |
|
May 8 |
revised |
Tests for spatial stationarity (homogeneity) added 3 characters in body |
|
May 8 |
comment |
lasso and cross-validation (theoretical results) This is a pretty generic question about generalization error and empirical risk minimization. |
|
Apr 27 |
accepted | Ripley's K Function and L Function for Point Patterns |
|
Apr 26 |
revised |
Ripley's K Function and L Function for Point Patterns added 125 characters in body |
|
Apr 26 |
asked | Ripley's K Function and L Function for Point Patterns |
|
Apr 26 |
asked | Tests for spatial stationarity (homogeneity) |
|
Apr 18 |
comment |
Errors in Variables and Deming's multivariate regression: Assumptions That's right..just linear regression |
|
Apr 8 |
comment |
Estimator bias without a closed form? Only the 'estimator' is computed iteratively. The theoretical part you mention is what is of interest to me. I need a starting point or example in this direction of computing expectations for an iterative update or some theoretical setting regardless of the iteration, that checks for the bias. It is even fine, if someone had a naive iterative algorithm for estimating the beta's in a linear regression ;) though we know that the OLS estimator is the same as MLE. But if there was an iterative estimation, how would the unbiasedness of $\hat{\beta}$ be verified? ;) |
|
Apr 8 |
comment |
Estimator bias without a closed form? Agreed. Will edit right now, to put in information about my estimator, followed by the distributional assumptions I make. Likewise, I have been wondering if you can put in a pointer to a paper with an iterative scenario where the expectation of the estimator is computed within a regression setting? I agree there are many such cases from 'REML' to Alternating Least Squares to M-estimation and so on and so forth. I would like a pointer to a paper that covers this scenario. |
|
Apr 8 |
revised |
Estimator bias without a closed form? added 15 characters in body; edited tags |
|
Apr 8 |
revised |
Estimator bias without a closed form? added 64 characters in body |
|
Apr 8 |
comment |
Errors in Variables and Deming's multivariate regression: Assumptions That's right @whuber. I meant independent variables, and so- edited it. |
|
Apr 8 |
revised |
Errors in Variables and Deming's multivariate regression: Assumptions added 2 characters in body |
|
Apr 8 |
revised |
Estimator bias without a closed form? added 22 characters in body |
|
Apr 8 |
revised |
Estimator bias without a closed form? deleted 116 characters in body |
|
Apr 8 |
comment |
Estimator bias without a closed form? @whuber, I edited/posed the problem am facing. Would you prefer that I put forth the distributional assumptions? I'd like to see your perspectives, as my issue is concerned with the lack of a closed-form. But, there may be a different view, on the problem of estimating the bias..that is not concerned with the lack of a closed form? |
|
Apr 8 |
revised |
Estimator bias without a closed form? deleted 22 characters in body |
