Resampling is taking a sample from a sample. Common uses are jackknifing (taking a subsample, eg all values but 1) & bootstrapping (sampling w/ replacement). These techniques can provide a robust estimate of a sampling distribution when it would be difficult or impossible to derive analytically.

learn more… | top users | synonyms

5
votes
2answers
32 views

what are the assumptions of permutation test?

It's often stated that permutation tests have no assumptions, however this is certainly not true. For example if my samples are somehow correlated, I can imagine that permuting their labels would not ...
7
votes
2answers
248 views

Best suggested textbooks on Bootstrap resampling?

I just wanted to ask which are in your opinion the best available books on bootstrap out there. By this I don't necessarily only mean the one written by its developers. Could you please indicate ...
1
vote
1answer
18 views

When do randomization methods outshine classical and non-parametric test

In which situations is it advantageous to use resampling methods as opposed to classical tests (if the data fit a certain distribution) or non-parametric tests? Is there a dataset/example that can ...
0
votes
0answers
28 views

How to use/interpret bootstrapping?

I am working on an stochastic optimization problem. Now I have come across the idea of using Monte Carlo sampling approach to solve it. I need the empirical distribution or the true distribution of ...
0
votes
0answers
22 views

Observational data, matching, and resampling. Is this method defensible?

I am interested in comments on the validity (or not) of the following method for propensity score analysis. I’ll simplify the scenario a bit for clarity. I have 1000 subjects that received the ...
0
votes
1answer
26 views

Bootstrap significance test

I'm using the bootstrap method to test my experiment results for significance. I have two sets (say A & B) of 50 grades, for which I want to test whether their means are significantly different. ...
0
votes
0answers
41 views

Quantifying the predictive ability of a model developed from a huge data set? (variation of bootstrapping?)

I have a statistical model with around 20 predictor variables, built on 90% of a dataset consisting of over 600k observations. The original developer held out 10% of the original dataset for the ...
0
votes
0answers
21 views

Fundamental Issues with Influence weighted resampling for bootstrapped predictions

I have a large database 1mill+ from which it is known that there are many influential points and outliers. I am interested in generating a series of predictions from subsets (1,000+) of the data and ...
3
votes
1answer
82 views

Bootstrapping the data to set up a prior

I am using a Gaussian model with a conjugate Normal-Inverse-Wishart (NIW) prior, as described here. The advantage of this approach is that the marginal likelihood $p(y)$, which is what I am interested ...
3
votes
1answer
67 views

Why use stratified cross validation? Why does this not damage variance related benefit?

I've been told that is beneficial to use stratified cross validation especially when response classes are unbalanced. If one purpose of cross-validation is to help account for the randomness of our ...
0
votes
1answer
30 views

Reservoir sampling with a computationally expensive weight function

I have a large dataset, and I want to obtain a small sample of it of size K, weighted by a function f(x) which is expensive to compute (I'm ok computing it O(K) times, but not too much more). Suppose ...
0
votes
1answer
31 views

sampling from confidence intervals

If we are provided a mean and (95%) confidence interval, is it possible to set up a system in which we draw random values that are outside the CIs 5% of the time? My intuition is that if one can ...
1
vote
1answer
52 views

parametric bootstrap for low sample sizes

I believe that this question is sufficiently different from previous related ones to warrant a new post. (I apologize if it has been answered already) I need to decide between various resampling ...
0
votes
0answers
19 views

Finding the fewest needed samples for a regression method

I have a regression method that I have applied with success to tens of thousands of well-observed objects. I'm looking to estimate how well it works on poorly observed objects. I selected 50 objects ...
7
votes
1answer
205 views

Is bootstrapping appropriate for this continuous data?

I'm a complete newbie :) I'm doing a study with a sample size of 10,000 from a population of about 745,000. Each sample represents a "percentage similarity". The great majority of the samples are ...
0
votes
0answers
31 views

How to improve results when using sampling in skewed binary classification?

I am using a data set with 18 features with True/False output (Related to mobile ad targeting). True values occurs only 0.4 % of the time. So, I have used sampling to keep the ratio of True and False ...
2
votes
0answers
74 views

Estimating means of correlated distributions with long tails

Suppose I have a relatively large number of samples (~1k) drawn from a series (~40) of increasingly long-tailed distributions (going from approximately normal to approximately log-normal). I want to ...
13
votes
2answers
2k views

resampling / simulation methods: monte carlo, bootstrapping, jackknifing, cross-validation, randomization tests, and permutation tests

I am trying to understand difference between different resampling methods (Monte Carlo simulation, parametric bootstrapping, non-parametric bootstrapping, jackknifing, cross-validation, randomization ...
1
vote
0answers
59 views

Oversampling with categorical variables

I would like to perform a combination of oversampling and undersampling in order to balance my dataset with roughly 4000 customers divided into two groups, where one of the groups have a proportion of ...
2
votes
1answer
42 views

Convenience sampling - Distribution forcing?

I am conducting some experiments on a data set that was collected by convenience. It is a data set based on historical data, most of which is not digitized. I know the exact distribution of the ...
1
vote
0answers
11 views

Correcting for multiple comparisons using simulated distribution

To check if unilateral pairs (defined below*) are coordinated (i.e. move together) in a flock of N coordinated individuals, we generate hypothetical (null) distribution of a certain focal parameter of ...
0
votes
0answers
54 views

Determining characteristics of sampling sets for EFA/CFA/SEM

Dividing sample data into several sets seems to be a common approach in statistics. This is especially evident in predictive modeling, where samples are traditionally divided into two sets, usually ...
0
votes
0answers
66 views

What's the safest way to resample to ensure equal class frequencies in training data?

I am working on a number of EEG data sets for binary classification. A good example of one is publicly available here. If you look at what Matthias Kaper did to classify that set, one thing that ...
4
votes
1answer
168 views

Why not always use bootstrap CIs?

I was wondering how bootstrap CIs (and BCa in barticular) perform on normally-distributed data. There seems to be lots of work examining their performance on various types of distributions, but could ...
1
vote
1answer
93 views

Logistic regression and discrepant sample sizes between 0/1 groups

I currently working in a multivariate logistic model but I have a problem regarding the sample size of my observations: -The "success" (1) event group has a sample size of 249 distinct observations - ...
1
vote
0answers
25 views

To find variance and covariance for a double sampling problem

A simple random sample of size $n=n_1 + n_2$ is drawn without replacement from a finite population of size $N$. Further a simple random sample of size $n_1$ is drawn without replacement from the first ...
2
votes
1answer
18 views

Doing comparisons from two null distributions

I have a situation where I have an observed statistic computed from data, and then I have approximated the null distribution by some sort of resampling. I have used this to calculate p-values for a ...
0
votes
0answers
31 views

Blocked Weighted Bootstrap

Bootstrap is a well known resampling method. But I want to know what is blocked weighted bootstrap sampling? Why we need this?
2
votes
0answers
52 views

What is the correct way of generating a p-value for a correlation with resampling?

I have a vector of gene expression values, across 20 patients. Each patient also has a glucose measure (a continuous numeric vector). I want to find how significant the real correlation value is ...
2
votes
0answers
20 views

Using Resampling to understand a large table

I have a data set that is very large. The attributes (columns) are several thousand. Some are sparse others are not. Some are ordinal, others interval, nominal or ratio. The row size is 10s of ...
1
vote
1answer
84 views

Addressing Non-response in a Convenience Sample

I am studying customer satisfaction in a large hierarchical organization. I plan to administer a voluntary survey to customers across the organization, and need to address non-response in my analysis. ...
0
votes
0answers
24 views

Estimating variance of prediction error in bootstrapped training sets with clustered data

I have C clusters with m elements each. I split the C clusters into a large training set D and a test set T. Hence, each element in D and T has m related elements, so its a cluster. I want to ...
2
votes
1answer
413 views

Optimal sampling strategy for EFA, CFA and SEM

I'm wondering what should be the optimal sampling strategy for my dissertation research. I have four data sources (two open source software projects meta-repositories and two global startup ...
0
votes
0answers
47 views

Analysis of forest inventory data - non-random samples

I apologise, this isn't a single question. It is more of a general problem on which I am working and am seeking guidance for how to proceed. I have been provided with an inventory dataset of plot ...
0
votes
1answer
174 views

What are RMSE SD and Rsquared SD metrics in resampling results using R package:caret?

I've been doing predictive modelling with R package caret. When resampling regression models, I get the traditional RMSE and Rsquared metrics, but also RMSE SD and ...
0
votes
0answers
51 views

Bootstrapping data with only sampling weights given

Suppose you only have these information from a sample data: $X_i$ and $w_i$, $i=1,...,N$, where $w_i$'s are the respective sampling weights(not integers). Is it possible to obtain a valid bootstrap ...
2
votes
0answers
58 views

Bias and variance estimation with boostrap

The Wikipedia article about Jacknife estimation of the bias and variance of an estimator $\theta$ includes the following formulas: Variance of $\theta$: $ \operatorname {Var}(\theta )=\sigma ...
2
votes
0answers
57 views

Statistical thought experiment (possibly Bayesian) about survey sampling and propensity scores

For some practical application I recently came about the following thought experiment. Can anybody help? Suppose we administer a survey A to measure a variable $Y$. The response probabilities may ...
1
vote
0answers
39 views

Bonferroni correction: control vs. groups?

I'm trying to understand how to set up a Bonferroni correction on several different groups and compare it to the control group. The groups and observations are as follows (with Group 0 being the ...
0
votes
2answers
66 views

Advice on resampling scheme for a small sample/ costly computation situation?

I am testing combinations of preprocessing activities on a small data set (n=48, p=30). The script generates 3200 different versions of the original data and measures how they perform in a ...
4
votes
1answer
166 views

Can Bootstrap Resampling be used to Calculate a Confidence Interval for the Variance of a Data Set?

I know that if you re-sample from a data set many times and calculate the mean each time, these means will follow a normal distribution (by the CLT). Thus, you can calculate a confidence interval on ...
2
votes
1answer
552 views

Bootstrapped p-value

I have a p-value that I generate via resampling. Resamples = 5000 Positive findings = 1000 positive findings P-value = 1000/5000 = 0.2 How can I compute the 95% confidence interval for this ...
0
votes
1answer
72 views

Selecting uncorrelated samples from a set of bulk data that contains correlated and dependent samples

i have a set of data that is generated by expensive computational model evaluations, on a total data set of 10000 samples in 40 dimensions. This sample data set is composed of different data sets, ...
2
votes
0answers
206 views

Sampling and resampling data in R

My problem this time concerns sampling size-related errors, resample-based confidence intervals and a possible way to control for this error. My dataset consists of 50 measurements of certain cranial ...
1
vote
0answers
75 views

Sufficient statistics and parametric bootstrapping

Does the resampling step of the bootstrap method require to have the sample entirely, or a sufficient statistic suffices? In general, the bootstrapping can be nonparametric or parametric. In the ...
3
votes
0answers
117 views

Bootstrapping fits to a small sample

I have a sample of experimentally measured survival times that are quite noisy and vary stochastically. The survival probability of these events (number of events with a survival time of t or more) is ...
2
votes
1answer
455 views

Required number of permutations for a permutation-based p-value

I need to calculate a permutation-based p-value with significance level $\alpha$, how many permutations do I need ? From the article "Permutation Tests for Studying Classifier Performance", page 5 : ...
0
votes
0answers
69 views

time series (asynchronous/variable) sample rate conversion

I am having a problen in preparing my data. The data is a GPS track recorded from a lengthy car-trip. This is recorded with a constant frequency - I have a datapoint every 200ms. Now I have ...
1
vote
1answer
72 views

How small can a subsample in bagging be before performance degrades severely?

If I want to perform bagging, would subsamples with sizes of 0.1% of the actual data be appropriate? The reason I want to do so is because my actual data set is very large in the tens of millions.
4
votes
1answer
219 views

Bootstrapping in R using the boot {boot} and Boot {car}

I'm trying my hand at resampling techniques with a dataset I have, and I think either I'm missing a conceptual point with bootstrapping, or I'm doing something incorrectly in ...