Id
stringlengths
1
6
PostTypeId
stringclasses
7 values
AcceptedAnswerId
stringlengths
1
6
ParentId
stringlengths
1
6
Score
stringlengths
1
4
ViewCount
stringlengths
1
7
Body
stringlengths
0
38.7k
Title
stringlengths
15
150
ContentLicense
stringclasses
3 values
FavoriteCount
stringclasses
3 values
CreationDate
stringlengths
23
23
LastActivityDate
stringlengths
23
23
LastEditDate
stringlengths
23
23
LastEditorUserId
stringlengths
1
6
OwnerUserId
stringlengths
1
6
Tags
list
7126
2
null
7103
1
null
Your distribution is parametric, and you should just store the parameters that are sufficient statistics, if you can identify them. That includes the distribution family. For a time series, you can take advantage of autocorrelation and store the parameters of the predictive distribution conditional on its previous valu...
null
CC BY-SA 2.5
null
2011-02-12T11:19:00.807
2011-02-12T11:26:39.763
2011-02-12T11:26:39.763
2456
2456
null
7127
2
null
7111
5
null
To add on Gilead's answer, they are computationally less expensive algorithms for truncated PCAs. NIPALS is indeed very popular, but I have had a lot of success with approximate methods that perform a succession of fits on partial data (what is often called PCA by random projection). This was discussed in a [metaoptimi...
null
CC BY-SA 4.0
null
2011-02-12T12:46:44.533
2020-02-05T16:42:23.473
2020-02-05T16:42:23.473
83252
1265
null
7128
1
null
null
5
610
### Study: I simulated some surfaces materials at audio and haptic level, and I asked subjects to evaluate on a 9-point Likert scale the degree of coherence between the two stimuli. For example there are stimuli with metal at auditory level and snow at haptic level, or wood both at auditory and haptic level. So it i...
Dissimilarity rating experiment: how to analyze the results
CC BY-SA 2.5
null
2011-02-12T15:22:27.697
2011-02-16T01:26:19.933
2011-02-15T01:10:26.023
183
4701
[ "rating" ]
7129
1
7130
null
5
45499
When presenting data using a percentage, is it a good thing to have decimal places, say 2 decimal places instead of rounding off to whole numbers? For example, instead of 23.43%, you round off to 23%. I am looking at this from the perspective of whether the 2 decimal places accuracy will make much difference since we a...
Should percentages be reported with decimal places?
CC BY-SA 3.0
null
2011-02-12T15:39:25.123
2023-05-17T01:52:02.650
2018-03-30T19:02:30.897
22228
546
[ "reporting", "percentage" ]
7130
2
null
7129
8
null
It depends on the size of the differences between classes. In most applications, saying the 73% prefer option A and 27% prefer option B is perfectly acceptable. But if you're dealing in an election where candidate X has 50.15% of votes and candidate Y has 49.86%, the decimal places are very much necessary. Of course, y...
null
CC BY-SA 2.5
null
2011-02-12T16:12:18.203
2011-02-12T16:12:18.203
null
null
666
null
7131
2
null
7054
1
null
In light of the clarifications in the comments to the question, I'd suggest using logistic regression with indicator variables for each region and a linear trend over year, and also consider adding an interaction between year and region to allow the linear trend to differ between regions. If you want something simpler,...
null
CC BY-SA 2.5
null
2011-02-12T16:39:30.410
2011-02-12T16:39:30.410
null
null
449
null
7132
1
7135
null
6
2654
When conducting a t-test we can always use a Welch approximation of the df, and let go of the assumption of equal variances. Is there something similar for (even one way) anova? (any R implementation of that?)
Is there an anova procedure that doesn't assume equal variance?
CC BY-SA 2.5
null
2011-02-12T16:46:03.520
2011-02-13T10:51:01.757
null
null
253
[ "r", "anova", "variance", "degrees-of-freedom" ]
7133
2
null
7128
1
null
ANOVA sounds reasonable to me — two-way ANOVA, to be specific, treating subjects as the blocking factor.
null
CC BY-SA 2.5
null
2011-02-12T16:46:10.710
2011-02-12T16:46:10.710
null
null
449
null
7134
1
16943
null
33
1003
Sometimes in reports I include a disclaimer about the p-values and other inferential statistics I've provided. I say that since the sample wasn't random, then such statistics would not strictly apply. My specific wording is usually given in a footnote: > "While, strictly speaking, inferential statistics are only...
Accommodating entrenched views of p-values
CC BY-SA 3.0
null
2011-02-12T17:12:53.863
2014-02-28T04:51:36.680
2011-10-13T03:32:05.563
919
2669
[ "inference", "p-value" ]
7135
2
null
7132
4
null
The latest version of [ez](http://cran.r-project.org/web/packages/ez/index.html) lets you pass a `white.adjust` argument to `car::Anova()`, which implements a correction for heteroscedasticity. See `?car::Anova()` for details.
null
CC BY-SA 2.5
null
2011-02-12T17:26:21.177
2011-02-12T17:26:21.177
null
null
364
null
7136
2
null
7134
13
null
The use of inferential statistics can be justified not only based on a population model, but also based on a randomization model. The latter does not make any assumptions about the way the sample has been obtained. In fact, Fisher was the one that suggested that the randomization model should be the basis for statistic...
null
CC BY-SA 2.5
null
2011-02-12T17:29:41.060
2011-02-12T17:29:41.060
null
null
1934
null
7138
2
null
7132
8
null
There is a function named `oneway.test()` in the base `stats` package, which implements Welch correction for a one-way ANOVA. Its use is similar to the standard `t.test()` function. It is also referred to as O'Brien transformation (Biometrics 40 (1984), 1079--1087) and might be applied with two or more independent samp...
null
CC BY-SA 2.5
null
2011-02-12T17:54:14.780
2011-02-13T10:51:01.757
2011-02-13T10:51:01.757
930
930
null
7139
1
null
null
4
1708
I came across this study as part of a mock exam paper and was confused to say the least. ### Context: The study investigates cognitive and behavioural factor related to the experience of anxiety in MRI scanners. Participants completed the following questionnaires 5 mins after the scan: - a measure of the frequency ...
Can causality be inferred in a study with an experience followed by two sets of measures
CC BY-SA 2.5
null
2011-02-12T18:47:04.910
2011-02-14T08:32:27.437
2011-02-13T10:46:39.673
930
null
[ "causality" ]
7140
2
null
7139
1
null
I think you're on the right track. Before drawing any conclusions about causality, I'd want to know... ...to what degree the 97 were representative of the 130, and the 130, to the population of interest. ...the magnitude of any relationships found as well as their statistical sig. (Whether correlation or anova was us...
null
CC BY-SA 2.5
null
2011-02-12T23:40:45.743
2011-02-12T23:40:45.743
null
null
2669
null
7141
1
null
null
7
224
Suppose I draw $n$ observations $X_1,X_2,\ldots,X_n$ independently from a distribution where $X_i \sim \mathcal{N}(\mu_i,\sigma^2)$, where the mean is assumed to be Lipschitz: $\left| \mu_i - \mu_{i+1}\right| \le \gamma,$ with $\gamma$ known. I want to test the null hypothesis: $$H_0: \mu_n = 0$$ against the local alt...
Location test under a bounded non-stationarity?
CC BY-SA 2.5
null
2011-02-13T03:57:04.277
2011-09-11T08:49:49.133
2011-02-15T03:25:37.087
795
795
[ "time-series", "hypothesis-testing", "stationarity" ]
7142
2
null
7139
1
null
### Design: The study is an observational design. Some may call it a correlational design, but I'm not a fan of such terminology because it can encourage the false assumption that the ability to draw a causal inference is related to the statistical test used. Following up on this point, ANOVA does not permit causal ...
null
CC BY-SA 2.5
null
2011-02-13T05:34:38.733
2011-02-13T05:34:38.733
null
null
183
null
7144
2
null
7111
14
null
What you're doing right now is close, but you need to make sure you multiply the eigenvectors of `(data . data.T) / lines` on the left by `data.T`, in order to get the eigenvectors of `(data.T . data) / lines`. This is sometimes called the "transpose trick". Here are some more details. Suppose you have a matrix $A$ tha...
null
CC BY-SA 3.0
null
2011-02-13T07:21:25.333
2011-06-07T21:15:04.847
2011-06-07T21:15:04.847
1106
1106
null
7146
1
7148
null
13
2784
Hi i am analaysing wind data for estimating energy from a wind turbine. I have taken 10 years of wind data and graphed a histogram; my second stage was to fit a Weibull distribution to the data. I used R with the package `lmom` to compute the Weibul shape and scale this is the code i used: ``` >library(lmom) wind....
Analysing wind data with R
CC BY-SA 2.5
null
2011-02-13T07:56:23.410
2011-04-04T15:12:40.560
2011-02-13T08:47:00.057
3178
3178
[ "r", "distributions" ]
7147
2
null
7110
11
null
I will add that in time series context it is usually assumed that data observed is a realisation of stochastic process. Hence in time series a lot of attention is given to properties of stochastic processes, such as stationarity, ergodicity, etc. In longitudinal context in my understanding data comes from usual sample...
null
CC BY-SA 2.5
null
2011-02-13T08:38:52.733
2011-02-13T08:38:52.733
null
null
2116
null
7148
2
null
7146
5
null
`lmom` function `pelwei` fits a three parameter Weibull distribution, with location, scale and shape parameters. `rweibull` generates random numbers for a two-parameter Weibull distribution. You need to subtract the location parameter `moments["zeta"]`. That should give a better fit, but it doesn't appear it will give ...
null
CC BY-SA 2.5
null
2011-02-13T09:51:27.640
2011-02-13T09:51:27.640
null
null
449
null
7149
2
null
6770
4
null
You can simply use the geometric distribution, which gives you the probability of a given number of Bernouilli trials before a success or in these case before the series is broken. The distribution is given as follow : P = (1-p)^(k) * p with p = .5 for a fair coin and k = 1 : 9 The idea is to compute the joint probabil...
null
CC BY-SA 2.5
null
2011-02-13T10:28:48.027
2011-02-13T11:02:14.957
2011-02-13T11:02:14.957
3108
3108
null
7150
2
null
7146
6
null
I recreated your plot with data from [http://hawaii.gov/dbedt/ert/winddata/krab0192.txt](http://hawaii.gov/dbedt/ert/winddata/krab0192.txt) (I took the 1200 measurements). I got a decent fit of the data, generally using your code: ``` library(lmom) daten <- read.delim("wind.txt") wind.avg <- na.omit(as.numeric(daten[,...
null
CC BY-SA 2.5
null
2011-02-13T11:21:23.203
2011-02-14T12:42:02.070
2011-02-14T12:42:02.070
1766
1766
null
7151
2
null
7128
2
null
It's hard to name an appropriate method without knowing the research question you're trying to answer. With that in mind, multidimensional scaling (MDS) takes measures of global dissimilarity between pairs of stimuli as input data: observers are asked to rate the similarity of two stimuli without being given explicit c...
null
CC BY-SA 2.5
null
2011-02-13T11:22:57.903
2011-02-13T12:34:08.807
2011-02-13T12:34:08.807
1909
1909
null
7152
1
7470
null
11
29836
How should you deal with a cell value in a contingency table that is equal to zero in statistical calculations? (Note that such a value can be structural, i.e., it must be zero by definition, or random, i.e., it could have been some other value, but zero was observed.)
How should you handle cell values equal to zero in a contingency table?
CC BY-SA 3.0
null
2011-02-13T13:14:55.147
2019-12-09T23:41:48.123
2013-12-08T03:41:59.073
7290
2956
[ "contingency-tables" ]
7153
1
7163
null
1
190
I would like to have a good idea on how to design clinical trials in oncology. In that issue, I am looking for a compact book that could give me a good overview, with the emphasis on statistical considerations. Would you have a recommendation for me? Thank you in advance, Marco
Book on designing clinical trials in oncology
CC BY-SA 2.5
null
2011-02-13T14:05:19.347
2011-02-13T23:51:13.627
2011-02-13T23:51:13.627
183
3019
[ "references", "experiment-design", "clinical-trials" ]
7155
1
7158
null
47
10555
People often talk about dealing with outliers in statistics. The thing that bothers me about this is that, as far as I can tell, the definition of an outlier is completely subjective. For example, if the true distribution of some random variable is very heavy-tailed or bimodal, any standard visualization or summary s...
Rigorous definition of an outlier?
CC BY-SA 2.5
null
2011-02-13T15:07:40.937
2013-08-27T22:26:05.173
null
null
1347
[ "outliers", "definition" ]
7156
1
23742
null
4
840
I have a dataset with yearly levels of corruption in a number of countries, as well as whether they changed their government that year. ``` year, corruption, change of president 2001, 5, 0 2002, 7, 1 2003, 8, 0 etc. ``` I want to test whether corruption is affected by a change in power (defined as the election of a ne...
Difference between two slopes
CC BY-SA 2.5
null
2011-02-13T16:00:25.593
2012-02-27T13:59:25.527
2011-02-13T22:06:14.257
2970
3182
[ "hypothesis-testing" ]
7157
2
null
7155
14
null
You are correct that removing outliers can look like a subjective exercise but that doesn't mean that it's wrong. The compulsive need to always have a rigorous mathematical reason for every decision regarding your data analysis is often just a thin veil of artificial rigour over what turns out to be a subjective exerc...
null
CC BY-SA 3.0
null
2011-02-13T16:03:53.543
2012-11-18T08:13:16.550
2012-11-18T08:13:16.550
601
601
null
7158
2
null
7155
26
null
As long as your data comes from a known distribution with known properties, you can rigorously define an outlier as an event that is too unlikely to have been generated by the observed process (if you consider "too unlikely" to be non-rigorous, then all hypothesis testing is). However, this approach is problematic on t...
null
CC BY-SA 3.0
null
2011-02-13T16:32:51.340
2013-08-27T20:01:56.297
2013-08-27T20:01:56.297
17230
198
null
7160
2
null
7152
18
null
Zeros in tables are sometimes classified as structural, i.e.zero by design or by definition, or as random, i.e. a possible value that was observed. In the case of a study where no instances were observed despite being possible, the question often comes up: What is the one-sided 95% confidence interval above zero? This ...
null
CC BY-SA 4.0
null
2011-02-13T17:16:00.367
2019-12-09T23:41:48.123
2019-12-09T23:41:48.123
102647
2129
null
7162
2
null
7155
6
null
Definition 1: As already mentioned, an outlier in a group of data reflecting the same process (say process A) is an observation (or a set of observations) that is unlikely to be a result of process A. This definition certainly involves an estimation of the likelihood function of the process A (hence a model) and sett...
null
CC BY-SA 3.0
null
2011-02-13T19:30:44.873
2013-08-27T20:00:52.887
2017-04-13T12:44:45.640
-1
223
null
7163
2
null
7153
2
null
I asked a very similar question several months back in this post: [Good Text on Clinical Trials](https://stats.stackexchange.com/questions/2770/good-text-on-clinical-trials). I decided to go with [Clinical Trials: A Methodologic Perspective](http://rads.stackoverflow.com/amzn/click/0471727814), by Steven Piantadosi I...
null
CC BY-SA 2.5
null
2011-02-13T19:34:36.337
2011-02-13T19:34:36.337
2017-04-13T12:44:48.803
-1
1118
null
7164
1
null
null
10
2335
I have a logistic regression model: $softmax(WX)$ where $W$ is my parameter matrix and $X$ is my input. I want a density function over the outputs of that model. Say I know that my $X$ are distributed according to some density $p$. From the [change of variables](http://en.wikipedia.org/wiki/Probability_density_function...
Change of variable with non bijective function
CC BY-SA 2.5
null
2011-02-13T20:15:29.043
2019-07-12T15:14:14.910
null
null
2860
[ "distributions", "probability" ]
7165
1
7168
null
12
649
What book is the most thorough treatment of fundamental concepts in statistics? I am not asking for a book on details of the methods of calculations and procedures, I am mainly interested in a book that thoroughly explains the foundational concepts ... an intuitive/illustrated/visual approach to the core ideas ... rat...
A resource on concepts underlying statistics, not the techniques used in applied stats
CC BY-SA 2.5
null
2011-02-13T20:24:38.900
2011-02-14T20:08:32.530
2011-02-13T23:28:51.173
159
1620
[ "references" ]
7166
1
7195
null
9
1045
a friend of mine has asked me to help him with predictive modelling of car traffic in a medium sized parking garage. The garage has its busy and easy days, its peak hours, dead hours opening hours (it is opened during 12 hours during weekdays and during 8 hours during weekends). The goal is to predict how many cars wi...
General approaches to model car traffic in a parking garage
CC BY-SA 2.5
null
2011-02-13T21:04:03.320
2011-04-13T10:34:55.570
2011-04-13T10:34:55.570
449
3184
[ "time-series", "multivariate-analysis", "predictive-models", "queueing" ]
7167
2
null
7146
4
null
Here's a recent post at SO on wind turbines. My answer on that link has three links that you might be interested in: [https://stackoverflow.com/questions/4843194/r-language-sorting-data-into-ranges-averaging-ignore-outliers/4844783#4844783](https://stackoverflow.com/questions/4843194/r-language-sorting-data-into-range...
null
CC BY-SA 2.5
null
2011-02-13T21:20:12.683
2011-02-14T01:24:19.100
2017-05-23T12:39:26.167
-1
2775
null
7168
2
null
7165
8
null
It's hard to know exactly what you're looking for based on your post. Maybe you can edit it to clarify a little. I will say that to really understand statistics well, then you'll need to learn some math. For fairly broad, low-level, introductory concepts, both - Gonick and Smith, A Cartoon Guide to Statistics, and - ...
null
CC BY-SA 2.5
null
2011-02-13T21:40:37.220
2011-02-13T22:37:46.510
2011-02-13T22:37:46.510
2970
2970
null
7169
2
null
7058
4
null
This sounds like a problem of feature selection, if this is the case, I think you want to compute the [mutual information](http://en.wikipedia.org/wiki/Mutual_information) between all subsets of features and the classification output. The subset with the highest mutual information will be the set of features that conta...
null
CC BY-SA 2.5
null
2011-02-13T22:04:12.390
2011-02-13T22:04:12.390
null
null
1913
null
7170
2
null
6753
1
null
Equi-depth histograms are a solution to the problem of [quantization](http://en.wikipedia.org/wiki/Quantization_%28signal_processing%29) (mapping continuous values to discrete values). For finding the best number of bins, I think it really depends on what you are trying to do with the histogram. In general I think it w...
null
CC BY-SA 2.5
null
2011-02-13T22:14:43.607
2011-02-13T22:14:43.607
null
null
1913
null
7171
2
null
7165
5
null
If you're interested in the philosophy of Statistics, you can't do much better than [Abelson's "Statistics As Principled Argument"](http://rads.stackoverflow.com/amzn/click/0805805281).
null
CC BY-SA 2.5
null
2011-02-13T22:17:14.337
2011-02-13T22:17:14.337
null
null
666
null
7173
1
7174
null
3
9104
I have a case-control study in which the cases are firms with health insurance and the controls are firms with no health insurance. I am studying the factors affecting enrolment in health insurance and was therefore using a logistic regression, which includes several covariates on firm characteristics that were measure...
How to decide between a logistic regression or conditional logistic regression?
CC BY-SA 2.5
null
2011-02-13T23:42:23.543
2017-08-01T03:12:36.997
2013-09-03T10:49:44.973
21599
834
[ "logistic", "survey", "clogit" ]
7174
2
null
7173
4
null
I don't agree that you sampled on the outcome, since you sampled on company and enrollment is your outcome. You may want to deal with the company as a random effect and the other features as fixed effects. So I am suggesting yet a third alternative: generalized mixed models. After clarification: If the outcome is compa...
null
CC BY-SA 2.5
null
2011-02-13T23:50:05.563
2011-02-16T21:06:49.663
2011-02-16T21:06:49.663
2129
2129
null
7175
1
7182
null
14
18329
I'm experimenting with classifying data into groups. I'm quite new to this topic, and trying to understand the output of some of the analysis. Using examples from [Quick-R](http://www.statmethods.net/advstats/cluster.html), several `R` packages are suggested. I have tried using two of these packages (`fpc` using the `k...
Understanding comparisons of clustering results
CC BY-SA 2.5
null
2011-02-14T00:21:17.413
2016-06-03T09:03:44.003
2011-02-20T21:59:36.737
2635
2635
[ "r", "clustering" ]
7176
1
7189
null
6
1953
Everything I've read about combining errors in quadrature when multiplying or dividing quantities with associated errors says that this works for "small error". What about large error? Say I want to compute A/B where A is +/- 1% and B is +/- 50%, can I still reasonably add the errors in quadrature?
Propagation of large errors
CC BY-SA 2.5
null
2011-02-14T00:28:54.207
2011-02-14T07:51:51.503
null
null
629
[ "error-propagation" ]
7177
2
null
7176
1
null
The formula for error propagation $\sigma_f^2 = \Sigma (\frac{\delta f}{\delta x} \sigma_x)^2$ works exactly for normally distributed errors and linear functions $f(x_1,x_2,...)$ Since (most) functions can be linearly approximated, the above also works for small errors. For large errors, a symmetric distribution of $x...
null
CC BY-SA 2.5
null
2011-02-14T02:10:59.663
2011-02-14T07:19:16.113
2011-02-14T07:19:16.113
449
198
null
7178
2
null
7176
1
null
The first problem with large errors is that the expected value of the multiplication or division of the uncertain values will not be the multiplication or the division of the expected values. So while it is true that $E[X+Y]=E[X]+E[Y]$ and $E[X-Y]=E[X]-E[Y]$, it would usually not be true to say $E[XY]=E[X]E[Y]$ or $E[...
null
CC BY-SA 2.5
null
2011-02-14T02:30:12.930
2011-02-14T02:30:12.930
null
null
2958
null
7179
1
7193
null
2
138
Sorry if this is similar to another question, but I'm still trying to learn data analysis and I don't know what to search for. What I'd like to do is take some information that a potential customer enters into a web form and figure out the probability that they'll become a customer... giving our sales team an idea of w...
Resources for beginners - how to determine probability of user action based on certain criteria?
CC BY-SA 2.5
null
2011-02-14T04:33:34.633
2011-02-14T09:37:18.190
null
null
3187
[ "dataset", "references" ]
7180
1
null
null
8
5596
I have a dataset of project case studies for a new type of research method for Government agencies to support decision making activities. My task is to develop an estimation method based on past experience for future projects for estimation purposes. My dataset is limited to 50 cases. I have 30+ (potential) predictor...
Multiple regression with small data sets
CC BY-SA 2.5
null
2011-02-14T05:32:01.177
2011-02-14T06:47:37.777
null
null
3189
[ "regression", "small-sample" ]
7181
2
null
7180
3
null
As you want to select a few predictors from your data set, I would suggest a simple linear regression with $L_1$ penalty or using the LASSO (penalized linear regression). Your case is suited for regression with [LASSO](http://en.wikipedia.org/wiki/Lasso_%28statistics%29#LASSO_method) penalty as your sample size, $n = 5...
null
CC BY-SA 2.5
null
2011-02-14T05:53:15.473
2011-02-14T06:47:37.777
2011-02-14T06:47:37.777
1307
1307
null
7182
2
null
7175
13
null
First let me tell you that I am not going to explain exactly all the measures here, but I am going to give you an idea about how to compare how good the clustering methods are (let's assume we are comparing 2 clustering methods with the same number of clusters). - For example the bigger the diameter of the cluster, th...
null
CC BY-SA 3.0
null
2011-02-14T06:03:07.077
2016-06-03T09:03:44.003
2016-06-03T09:03:44.003
99963
1808
null
7185
1
7199
null
15
12435
Can anyone tell me the difference between using `aov()` and `lme()` for analyzing longitudinal data and how to interpret results from these two methods? Below, I analyze the same dataset using `aov()` and `lme()` and got 2 different results. With `aov()`, I got a significant result in the time by treatment interaction,...
What is the difference between using aov() and lme() in analyzing a longitudinal dataset?
CC BY-SA 3.0
null
2011-02-14T06:46:27.827
2012-09-07T02:47:09.103
2012-09-07T02:47:09.103
7290
1663
[ "r", "mixed-model", "repeated-measures", "panel-data" ]
7186
2
null
3
9
null
First of all let me tell you that in my opinion the best tool of all by far is R, which has tons of libraries and utilities I am not going to enumerate here. Let me expand the discussion about weka There is a library for R, which is called RWeka, which you can easily install in R, and use many of the functionalities fr...
null
CC BY-SA 4.0
null
2011-02-14T07:39:27.737
2022-11-27T23:33:13.540
2022-11-27T23:33:13.540
11887
1808
null
7188
2
null
7115
2
null
Lets suppose we can calculate the distance from one noun to another in the following way. Use the Worldnet (which I guess you know), and utilize a function that exists, but you can build it yourself, that counts for how many points of the taxonomy of words you need to get from one word to another (for example from cat ...
null
CC BY-SA 2.5
null
2011-02-14T07:51:24.210
2011-02-14T07:51:24.210
null
null
1808
null
7189
2
null
7176
5
null
For large error, the standard error of $A/B$ depends on the distributions of $A$ and $B$, not just on their standard errors. The distribution of $A/B$ is known as a [ratio distribution](http://en.wikipedia.org/wiki/Ratio_distribution), but which ratio distribution depends on the distributions of $A$ and $B$. If we assu...
null
CC BY-SA 2.5
null
2011-02-14T07:51:51.503
2011-02-14T07:51:51.503
null
null
449
null
7190
1
null
null
3
285
I have a set of integer that I am trying to see use different methods to categorize them into four groups, and the 2x2 table for the outcome of the 2 methods is displayed as below: ``` method_B method_A 0 1 2 3 0 182 11 0 0 1 41 127 2 0 2 0 1...
Statistical test for difference in categorization method
CC BY-SA 2.5
null
2011-02-14T08:16:36.003
2011-02-14T08:35:26.457
2011-02-14T08:18:23.990
null
588
[ "categorical-data", "chi-squared-test" ]
7191
2
null
7139
1
null
Here is another question from this websites that says among other things the following: What can a statistical model say about causation? This led to his motto: NO CAUSATION WITHOUT MANIPULATION which emphasized the importance of restrictions around experiments that consider causation. Andrew Gelman makes a similar poi...
null
CC BY-SA 2.5
null
2011-02-14T08:32:27.437
2011-02-14T08:32:27.437
2017-04-13T12:44:53.777
-1
1808
null
7192
2
null
7190
2
null
I don't think either chi-squared test or Fisher's exact test is useful here. It's pretty obvious that the results from method A and from method B aren't independent of each other. A statistic to quantify how well the two methods agree is more useful. The obvious choice is [Cohen's kappa](http://en.wikipedia.org/wiki/Co...
null
CC BY-SA 2.5
null
2011-02-14T08:35:26.457
2011-02-14T08:35:26.457
null
null
449
null
7193
2
null
7179
3
null
Your problem can be described as "binary classification problem" with the response variable "become_customer" $\in$ {yes,no}. As far as see, your next steps should be: - create a tabular dataset with one row = one visitor and columns = predictors + response variable. Predictor variables reflect the input into the form...
null
CC BY-SA 2.5
null
2011-02-14T09:37:18.190
2011-02-14T09:37:18.190
null
null
264
null
7194
2
null
7165
1
null
I like Kennedy's Guide to Econometrics, which treats every topic on three levels, the first of which is a non-technical description, in so far as this is possible.
null
CC BY-SA 2.5
null
2011-02-14T11:08:24.803
2011-02-14T11:08:24.803
null
null
1766
null
7195
2
null
7166
6
null
The field that is relevant to the problem is the [Queuing theory](http://en.wikipedia.org/wiki/Queueing_model), a particular sub-field is a [Birth-death](http://en.wikipedia.org/wiki/Birth-death_process) processes. An article that in my opinion is helpful to your task is R.C. Larson and K.Satsunama (2010) [Congestion P...
null
CC BY-SA 2.5
null
2011-02-14T11:47:39.873
2011-02-14T11:47:39.873
null
null
2645
null
7196
1
7203
null
3
234
The data I want to analyze consist of set of votes similar to the voting system here on stackexchange. Votes are binary, i.e, items can receive up- or down-votes. The data have been gathered in an A/B controlled experiment. I want to compare the control group to the treatment group according to some gold standard. Tha...
Appropriate test to compare online voting data
CC BY-SA 2.5
null
2011-02-14T13:47:16.960
2011-02-14T16:51:13.537
2011-02-14T14:48:20.330
null
3191
[ "hypothesis-testing" ]
7197
1
null
null
4
1887
I'm looking for sample R code, or pointers to sample R code for the following. (Gentle) critique of the approach would also be appreciated. I'm not a statistician and I'm pretty new to R. I have duration of hospitalization ("length of stay"=LOS) data for 2,000 patients from my hospital and 50,000 patients from a comp...
Calculating confidence interval for average hospital length of stay, case-mix adjusted, in R
CC BY-SA 3.0
null
2011-02-14T14:36:39.887
2011-06-26T21:39:02.970
2011-06-26T13:44:06.477
null
690
[ "r", "confidence-interval", "bootstrap" ]
7198
2
null
7185
0
null
It appears to me you have multiple measures for each id at each time. You need to aggregate these for the aov because it unfairly increases power in that analysis. I'm not saying doing the aggregate will make the outcomes the same but it should make them more similar. ``` dat.agg <- aggregate(UOP.kg ~ time + treat + ...
null
CC BY-SA 2.5
null
2011-02-14T15:10:27.500
2011-02-14T15:10:27.500
null
null
601
null
7199
2
null
7185
20
null
Based on your description, it appears that you have a repeated-measures model with a single treatment factor. Since I do not have access to the dataset (`raw3.42`), I will use the Orthodont data from the `nlme` package to illustrate what is going on here. The data structure is the same (repeated measurements for two di...
null
CC BY-SA 2.5
null
2011-02-14T15:35:47.063
2011-02-14T15:35:47.063
null
null
1934
null
7200
1
7206
null
24
7340
I know that an easy to handle formula for the CDF of a normal distribution is somewhat missing, due to the complicated error function in it. However, I wonder if there is a a nice formula for $N(c_{-} \leq x < c_{+}| \mu, \sigma^2)$. Or what the "state of the art" approximation for this problem might be.
Evaluate definite interval of normal distribution
CC BY-SA 2.5
null
2011-02-14T15:43:39.727
2022-05-21T11:21:16.657
null
null
2860
[ "normal-distribution", "approximation" ]
7201
2
null
7155
9
null
I don't think it is possible to define an outlier without assuming a model of the underlying process giving rise to the data. Without such a model we have no frame of reference to decide whether the data are anomalous or "wrong". The definition of an outlier that I have found useful is that an outlier is an observati...
null
CC BY-SA 2.5
null
2011-02-14T16:36:53.473
2011-02-14T16:36:53.473
null
null
887
null
7202
1
null
null
2
7239
### Context: I have "pooled data" with time and cross section dimensions. It is unbalanced data without a full range of time observations for each cross section of observations. It is kind of like unbalanced panel data? I wish to proceed with the limited number of observations without losing any data. ### Que...
Regression with Pooled data in SPSS
CC BY-SA 2.5
null
2011-02-14T16:40:18.977
2012-04-13T05:03:47.760
2011-02-15T11:20:15.173
183
null
[ "regression", "spss", "panel-data" ]
7203
2
null
7196
2
null
If you know what the correct vote should be, you could start with a simple chi-square or KS test, testing for the number of votes misclassified for each group (treatment and control) as compared to the gold standard. -Ralph Winters
null
CC BY-SA 2.5
null
2011-02-14T16:51:13.537
2011-02-14T16:51:13.537
null
null
3489
null
7205
2
null
7202
1
null
You should be able to weight by duration of observation, perhaps by simple division, in essence turning data into an annualized rates. There will be issues relating to whether persons (or other units of analysis) have systematically higher or lower rates in the early period of observation or enrollment.
null
CC BY-SA 2.5
null
2011-02-14T16:58:27.340
2011-02-14T16:58:27.340
null
null
2129
null
7206
2
null
7200
44
null
It depends on exactly what you are looking for. Below are some brief details and references. Much of the literature for approximations centers around the function $$ Q(x) = \int_x^\infty \frac{1}{\sqrt{2\pi}} e^{-\frac{u^2}{2}} \, \mathrm{d}u $$ for $x > 0$. This is because the function you provided can be decomposed a...
null
CC BY-SA 4.0
null
2011-02-14T17:01:06.277
2022-05-21T11:21:16.657
2022-05-21T11:21:16.657
79696
2970
null
7207
1
7210
null
250
118263
I understand the formal differences between them, what I want to know is when it is more relevant to use one vs. the other. - Do they always provide complementary insight about the performance of a given classification/detection system? - When is it reasonable to provide them both, say, in a paper? instead of just ...
ROC vs precision-and-recall curves
CC BY-SA 3.0
null
2011-02-14T17:10:17.143
2022-02-23T15:40:56.883
2015-03-09T12:32:29.693
2798
2798
[ "machine-learning", "roc", "precision-recall" ]
7208
1
7338
null
8
2393
I am using [Cohen's Kappa](http://en.wikipedia.org/wiki/Cohen%27s_kappa) to calculate the inter-agreement between two judges. It is calculated as: $ \frac{P(A) - P(E)}{1 - P(E)} $ where $P(A)$ is the proportion of agreement and $P(E)$ the probability of agreement by chance. Now for the following dataset, I get the exp...
Can one use Cohen's Kappa for two judgements only?
CC BY-SA 2.5
null
2011-02-14T17:16:52.580
2011-02-18T01:26:44.263
2011-02-17T22:01:30.213
930
1205
[ "reliability", "information-retrieval" ]
7209
1
null
null
5
945
My aim is to predict quarterly customer-default probabilities: I have data on ~ 2 million individuals, who default on average with a probability of ~ 0.3 percent. Therefore I am thinking about undersampling the majority class (non-defaults) to save computing-time (kernel methods can be quite costly, I know about correc...
How much undersampling should be done?
CC BY-SA 2.5
null
2011-02-14T17:40:19.597
2011-02-14T23:36:54.893
2011-02-14T18:40:11.043
null
2549
[ "sampling" ]
7210
2
null
7207
331
null
The key difference is that ROC curves will be the same no matter what the baseline probability is, but PR curves may be more useful in practice for needle-in-haystack type problems or problems where the "positive" class is more interesting than the negative class. To show this, first let's start with a very nice way to...
null
CC BY-SA 3.0
null
2011-02-14T18:11:52.737
2013-06-05T17:48:31.693
2013-06-05T17:48:31.693
7290
1347
null
7211
1
null
null
12
12708
Can you recommend a text mining package in R that can be used against large volumes of data? Secondly, is there a GUI available for any of the text mining packages in R? Thirdly, is there another open source text mining program that is easy and intuitive to use?
What are the text-mining packages for R and are there other open source text-mining programs?
CC BY-SA 2.5
null
2011-02-14T19:29:14.993
2016-04-02T16:53:50.907
2011-02-14T23:38:31.837
null
null
[ "r", "text-mining" ]
7212
2
null
7211
4
null
See the [tm package](http://tm.r-forge.r-project.org/) and [this presentation](http://www.rinfinance.com/agenda/2010/Theussl+Feinerer+Hornik.pdf) by [Stefan Theussl](http://statmath.wu-wien.ac.at/~theussl/) given at the 2010 [R/Finance conference](http://www.rinfinance.com).
null
CC BY-SA 2.5
null
2011-02-14T19:35:54.927
2011-02-14T19:35:54.927
null
null
1657
null
7213
2
null
7165
2
null
I think Harvey Motulsky's Intuitive Biostatistics is pretty good for non-mathematical "intuitive" explanations to basic statistical methods most commonly employed in the biological and medical sciences.
null
CC BY-SA 2.5
null
2011-02-14T20:08:32.530
2011-02-14T20:08:32.530
null
null
3183
null
7214
2
null
7211
17
null
Please see the [CRAN Task View on Natural Language Processing](http://cran.r-project.org/web/views/NaturalLanguageProcessing.html) which includes, among others, the [tm](http://cran.r-project.org/package=tm) package already mentioned by Josh.
null
CC BY-SA 2.5
null
2011-02-14T20:50:04.587
2011-02-14T20:50:04.587
null
null
334
null
7215
6
null
null
0
null
I have been one of the pro tem moderators and I would like to continue my mission as a regular mod. So, if you like what I was doing here (compulsive edits, protecting wide scope, journal club), vote for me!
null
CC BY-SA 2.5
null
2011-02-14T21:19:12.923
2011-02-14T21:19:12.923
2011-02-14T21:19:12.923
null
null
null
7216
2
null
7211
3
null
Sure, RapidMiner with the text mining extension. There are many videos that show how it is done.
null
CC BY-SA 2.5
null
2011-02-14T21:24:14.247
2011-02-14T21:24:14.247
null
null
null
null
7217
2
null
7211
1
null
GATE is very comprehensive. It also allows you to work with different languages and has an ontology editor.
null
CC BY-SA 3.0
null
2011-02-14T21:44:54.347
2011-11-21T22:20:49.540
2011-11-21T22:20:49.540
930
3489
null
7218
1
7221
null
10
3170
Suppose I have a large population of data points $(x,y)$ and that the Pearson's correlation is $$\textrm{corr}(X,Y) = \rho$$ What can I reasonably say about the correlation I expect to observe in a sample of size $n$? If the sample correlation is $\rho_s$, roughly what is the spread is $\rho_s$? Is $\rho_s$ biased? ...
Distribution of sample correlation
CC BY-SA 2.5
null
2011-02-14T22:05:32.307
2020-11-29T11:41:58.980
2020-11-29T11:41:58.980
11887
2665
[ "distributions", "correlation" ]
7219
2
null
6
197
null
The biggest difference I see between the communities is that statistics emphasizes inference, whereas machine learning emphasized prediction. When you do statistics, you want to infer the process by which data you have was generated. When you do machine learning, you want to know how you can predict what future data ...
null
CC BY-SA 2.5
null
2011-02-14T22:09:41.657
2011-02-14T22:09:41.657
null
null
1347
null
7220
2
null
770
4
null
Depends on what you mean. I've gotten deeper and deeper into statistics based on exposure to machine learning. (I had previously been more of a general AI guy, and hadn't had good experience with statistics, but gained greater understanding and appreciation of statistics as time has gone on.) So it's certainly a useful...
null
CC BY-SA 2.5
null
2011-02-14T22:29:44.980
2011-02-14T23:04:02.267
2011-02-14T23:04:02.267
1764
1764
null
7221
2
null
7218
10
null
To quote the [Wikipedia article on the Fisher transformation](http://en.wikipedia.org/wiki/Fisher_transformation) : If $(X, Y)$ has a bivariate normal distribution, and if the $(X_i, Y_i)$ pairs used to form the sample correlation coefficient $r$ are independent for $i = 1, \ldots, n,$ then $$z = {1 \over 2}\ln{1+r \ov...
null
CC BY-SA 2.5
null
2011-02-14T22:34:14.100
2011-02-15T08:55:41.067
2011-02-15T08:55:41.067
449
449
null
7222
2
null
7209
2
null
In a first approximation, 1:1 is a good proportion, but: - Some methods are more vulnerable to unequal classes, some are less -- plain decision tree will almost always vote for a much larger class, 1-NN will be not affected at all. It is a good idea to check this out (in literature or by asking here) in context of y...
null
CC BY-SA 2.5
null
2011-02-14T23:36:54.893
2011-02-14T23:36:54.893
null
null
null
null
7223
1
null
null
12
1833
I would like to pose this question in two parts. Both deal with a generalized linear model, but the first deals with model selection and the other deals with regularization. Background: I utilize GLMs (linear, logistic, gamma regression) models for both prediction and for description. When I refer to the "normal things...
GLM after model selection or regularization
CC BY-SA 4.0
null
2011-02-14T23:51:15.593
2018-06-20T09:40:15.533
2018-06-20T09:40:15.533
128677
2040
[ "regression", "model-selection", "regularization" ]
7224
1
null
null
110
14811
I'm working on a little project involving the faces of twitter users via their profile pictures. A problem I've encountered is that after I filter out all but the images that are clear portrait photos, a small but significant percentage of twitter users use a picture of Justin Bieber as their profile picture. In order ...
Detecting a given face in a database of facial images
CC BY-SA 2.5
null
2011-02-14T22:41:09.187
2012-02-19T18:09:33.437
2011-02-15T15:22:25.870
null
21466
[ "machine-learning", "clustering", "image-processing" ]
7225
1
null
null
20
10320
I am using the R package penalized to obtain shrunken estimates of coefficients for a dataset where I have lots of predictors and little knowledge of which ones are important. After I've picked tuning parameters L1 and L2 and I'm satisfied with my coefficients, is there a statistically sound way to summarize the model ...
Estimating R-squared and statistical significance from penalized regression model
CC BY-SA 2.5
null
2011-02-15T00:38:52.977
2018-04-05T13:56:12.613
2020-06-11T14:32:37.003
-1
36
[ "regression", "lasso", "stepwise-regression", "ridge-regression" ]
7226
2
null
7197
1
null
Given that you're comparing n groups ($n > 2$) and that your count data are, as you say, likely non-normally distributed, I wonder if perhaps applying Kruskal–Wallis analysis of variance to your data might be an option.
null
CC BY-SA 3.0
null
2011-02-15T00:58:32.400
2011-06-26T21:39:02.970
2011-06-26T21:39:02.970
930
2730
null
7227
2
null
7224
49
null
A better idea might be to trash all images that appear in the feed of more than one user - no recognition needed.
null
CC BY-SA 2.5
null
2011-02-15T01:03:52.237
2011-02-15T01:03:52.237
null
null
3203
null
7228
2
null
7128
0
null
A few thoughts - you might want to rearrange your data frame either in wide (rows are cases and columns are ratings for stimuli pairs) or long format (rows are case by stimuli pair combinations). Then you could run mixed models lme4 or perhaps aov (see here for info on different ANOVAs). - You might find useful the ...
null
CC BY-SA 2.5
null
2011-02-15T01:14:20.777
2011-02-16T01:26:19.933
2011-02-16T01:26:19.933
183
183
null
7229
2
null
7224
16
null
I have a feeling that [http://www.tineye.com/commercial_api](http://www.tineye.com/commercial_api) may be the solution here. Simply throw the Twitter profile image to Tineye, see if it returns images (and associated URLs) that can clearly be identified (or automatically scored using simple word-count logic) as being re...
null
CC BY-SA 2.5
null
2011-02-15T01:19:42.080
2011-02-15T01:19:42.080
null
null
3205
null
7231
2
null
7224
2
null
You could try [locality sensitive hashing](http://en.wikipedia.org/wiki/Locality_sensitive_hashing).
null
CC BY-SA 2.5
null
2011-02-15T02:28:19.450
2011-02-15T02:28:19.450
null
null
900
null
7232
2
null
7224
12
null
Since you are able to filter to only those that are clear portrait photos, I'm assuming you have some method of feature generation to transform the raw images into features that are useful for machine learning purposes. If that's true, you could try to train a classification algorithm (there are lots of them: neural n...
null
CC BY-SA 2.5
null
2011-02-15T02:52:31.463
2011-02-15T02:52:31.463
null
null
2485
null
7233
1
7243
null
1
1898
How to interpret residuals in random and fixed effects models? I am a medical doctor. Kindly explain in simple words. Updated Here is an example: ``` Weight (Fixed) Weight (Random) Residual (Fixed) Residual (Random) Relative weight Relative weight Std Residual Std Residual 32.27 ...
Interpret residuals in random and fixed effects models in meta-analysis
CC BY-SA 3.0
null
2011-02-15T03:05:32.217
2018-05-26T19:37:57.020
2018-01-30T18:08:43.177
101426
2956
[ "residuals", "meta-analysis" ]
7235
2
null
7223
6
null
You might check out David Freedman's paper, "[A Note on Screening Regression Equations.](http://www.psych.umn.edu/faculty/waller/classes/mult11/readings/freedman83.pdf)" (ungated) Using completely uncorrelated data in a simulation, he shows that, if there are many predictors relative to the number of observations, the...
null
CC BY-SA 2.5
null
2011-02-15T03:40:52.463
2011-02-15T03:40:52.463
null
null
401
null
7236
1
null
null
6
1998
I've got a dataset similar to this: ``` pat_id epis Care Type 1 1722650 Acute Care 1 1723120 Rehabilitation care 2 1584309 Acute Care 2 1585705 Rehabilitation care 3 1726487 GEM 3 1664031 Acute Care 3 1726488 Acute Care 3 1726489 Rehabilitation care ``` Each patient ha...
Multistage sampling in R
CC BY-SA 3.0
null
2011-02-15T04:22:10.493
2011-06-17T17:37:47.020
2011-06-17T17:37:47.020
null
null
[ "r", "sampling" ]
7238
1
null
null
3
2169
I have two vectors of matching lengths. They are readings from two different sensors (one is from a smartphone and the other is from a wiimote) of the same hand movement. I am trying to find the time offset between them to synchronise the readings for further processing. The readings I get are of the format (Time(ms) V...
MATLAB interpretation of xcorr2 cross correlation function
CC BY-SA 3.0
null
2011-02-15T05:53:31.987
2013-08-16T13:05:11.887
2012-10-19T09:34:45.953
null
null
[ "data-visualization", "correlation", "matlab", "interpretation", "cross-correlation" ]
7239
1
7264
null
4
5943
I am fitting a simple linear regression model with R & JAGS: ``` model { for (i in 1:length(ri)){ ri[i] ~ dnorm(mu[i],tau) mu[i] <- alpha + b.vr*vr[i] + b.ir*ir[i] } #posterior predictive distribution for (j in 1:length(pvr)){ pmu[j] <- alpha + b.vr*pvr[j] + b.ir*pir[j] ...
JAGS: posterior predictive distribution
CC BY-SA 2.5
null
2011-02-15T08:40:40.117
2011-02-18T21:58:49.337
2011-02-15T18:23:17.730
8
1443
[ "markov-chain-montecarlo", "bugs", "gibbs", "jags" ]