Questions tagged [computational-statistics]

Refers to the interface of statistics and computing; the use of algorithms and software for statistical purposes.

Filter by
Sorted by
Tagged with
1
vote
1answer
39 views

Computation of generalized least squares solutions of large sparse systems

Suppose $X$ and $\Omega$ are large sparse matrices, with $\Omega$ symmetric positive definite (but not diagonal), and $y$ is a vector. I want to find the generalized least squares solution: $$\hat\...
0
votes
0answers
14 views

Do I need to calculate proposal probability using random walk on exponential space?

I know I can use symmetric random walk to avoid computing the proposal probability. However, I'm not sure if doing symmetric random walk on transformed space should also cancel the proposal ...
1
vote
0answers
20 views

Computing KL Divergence for distributions over sets

I have a distribution over a set of (hundreds of) discrete terms, and I'd like to describe the difference between I see a couple of options, and none seems really attractive: Take the KL divergence ...
2
votes
4answers
169 views

which is the meaning of scatterplot between a pair of 2 consecutive pseudo random numbers with respect to the independence of the sequence?

Pseudo random number generators should give as output random sequences u1, u2, ... that are mutually independent and identically distribuited (iid). Since testing for independence is not easy, the ...
0
votes
1answer
96 views

Correlation Matrix or Covariance Matrix in PCA [duplicate]

I have 4 metrics, three of them measured on the scale 0 to 1, and one measured on the scale 0 to 6. When I stored my data, I converted the fourth one by dividing it by 10, so that I can get values ...
1
vote
0answers
103 views

Dirichlet Process Concentration Parameter - collapse at zero

Background I've implemented the blocked gibbs sampler for sampling from the posterior of a dirichlet process mixture model as described on p.552 of Bayesian Data Analysis, placing a Gamma prior on the ...
1
vote
0answers
70 views

How to confirm that the differences between clusters are statistically significant? [closed]

I have a DataFrame of page navigation behavior from visits to an e-commerce website. My independent variable is Revenue, which is simply binary (1 for Revenue ...
0
votes
1answer
44 views

Changepoint detection and confidence score

I am familiar with Bayes Theorem and hypothesis testing, but not much above that. However, I have a problem that I cannot seem to formulate in frequentist terms - to my knowledge. From what I know ...
2
votes
1answer
76 views

Detecting extrema under uncertainty?

An old version of this question was poorly articulated. Here is another go: I have fifty objects. With a different, independent, unbiased scale for each object, I measure their weights 100 times each ...
1
vote
0answers
73 views

Correlation between two 3d arrays [closed]

I know that a is related to b is related to c. I have data for two years: In 2018, d=10. In 2019, d=0. I would like to know the correlation between a, b and c, for both d=0 and d=10 in order to ...
2
votes
1answer
273 views

LDA/NMF Topic Modeling vs Topic Modeling using “skip gram” approach

I am having a little friendly debate with my coworker on how to properly/optimally do topic modeling. I am just using the regular traditional nmf/lda approach and he decided to do it using "skip grams"...
0
votes
0answers
200 views

Ordinal multinomial logistic regression on one-hot encoded data

I have a task I am unable to tackle by principle. I'm working on survey data for one of our clients such that my design matrix is made of one-hot vectors with 15 features (originally 3 variables with ...
0
votes
0answers
121 views

Random-effects-meta-analysis-simulation: zero-estimates for tau^2

I am working on simulating a random-effects-model for comparison of the DerSimonian-Laird-method vs. Hartung-Knapp-Sidik-Jonkman-method in R. To do so, I chose different combinations of mu (true ...
5
votes
1answer
63 views

Stable and efficient computation of binomial expectations

Suppose we want to compute the expected value of some function $f(X)$ where $X \sim \text{Bin}(n,\theta)$. Taking $\mathbf{f} = (f_0,...,f_n)$ to be the function values over all possible outcomes of ...
0
votes
1answer
43 views

Binary classification with binary predictors

I have a dataset with a binary response variable and 45 binary predictors. Which possible classification algorithms can I use? And how can i perform feature selection when most methods like best ...
1
vote
0answers
79 views

fitting high dim copula to residuals of a garch model very slow in R [closed]

I'm looking for some help on understanding on the fitting procedure of a normal (or any other for that matter) copula in R. My main goal is to either improve computational speed, or revise my strategy....
3
votes
1answer
908 views

Find true negatives in a confusion matrix

I'm trying to find the True negative in a confusion matrix, I have computed successfully from scratch the precision and recall/sensibility, now i need to compute the accuracy and specificity. This is ...
0
votes
1answer
89 views

Statistics: what kind of test to use to test this hypothesis?

I am unsure what kind of a hypothesis test to use in my case: So we are filling 500 ml bottles with water; It's a known fact that before a max of 3% of all the bottles could be under-filled; Now I ...
1
vote
1answer
112 views

Computationally verifying the equivalence of ridge regression estimates and Bayesian regression estimates

I'm trying to show that the numerical estimates of ridge regression's parameter estimates are the same as the MAP parameter estimates of a Bayesian regression model with normal prior distributions. So ...
1
vote
0answers
48 views

Error bars of Monte Carlo expectation with correlated samples

I will try to phrase the question in a general way, then give my specific case as an example. Suppose I want to evaluate $Q = \mathbb E \left[ f\left(X, Y \right) \right]$ where $X$ and $Y$ are ...
0
votes
1answer
36 views

How to test the difference of two linear regression slopes with 2 independent and one dependent variable

I am trying to determine if the CO2 emissions growth rate of developing countries is higher than the growth rate of developed countries. So essentially I need to compare two linear regression slopes ...
1
vote
0answers
17 views

create joint prob distribution or empirical relation for two variables

There are two variables, X1 and X2. The experimental study shows that they are highly correlated. Are there any reliable ways to create an empirical mapping(or equation) between X1 and X2. Assuming ...
1
vote
0answers
22 views

What does it mean when Risk function turns out to be a number?

I have a statistical decision making theory problem.I have to calculate the Risk Function for each of 4 decision rules.However,it turns out that the fourth Risk function is not a function of θ and it ...
4
votes
1answer
4k views

Why median is NA for some of the group outcomes in survival analysis?

I'm trying to do survival analysis using the Followup information, patient_vital_status and the expression of gene. I'm using like below: ...
0
votes
1answer
3k views

Use Shapley Values for explaining whole Data Frame instead of a Single prediction [closed]

I am working on a Machine Learning model. One of the requests is to explain the models 'decisions' to the business. Therefore I am using Shapley Values (Game Theory). I found an interesting example ...
6
votes
1answer
205 views

Does Approximate Bayesian Computation (ABC) follow the Likelihood Principle?

I know that ABC is commonly used when the likelihood is intractable, so likelihood principle is not an interest in that case. But, I am curious whether the ABC satisfies the likelihood principle when ...
1
vote
0answers
41 views

What is the process for detecting a 0.2 gallon/hour leak in an underground petrol tank with an accuracy of 95% or more?

Current EPA regulation stipulates that SIR (Statistical Inventory Reconciliation) vendors must be able to detect a leak of 0.2 gallons/hour with a probability of 0.95 or higher, and a false alarm rate ...
1
vote
0answers
265 views

Using R to maximize a two parameter Weibull model via multivariate extension of Newton-Raphson method

I am just getting back into using R for the first time in a while, and wrote some code to perform the aforementioned task in the title. I was wondering if anyone could take a look at it and see if ...
5
votes
1answer
1k views

Which is more numerically stable for OLS: pinv vs QR

If I am doing standard OLS and want to calculate beta values (OLS estimators), which of the following is the more numerically stable method? And why? Assuming that the columns of $X$ are already mean-...
0
votes
1answer
40 views

A simple concept question regarding Bootstrapping

I am having a difficulty understanding whether I can use the bootstrapping method for prediction. First off, my data is as follows. where personal Income is the dependent variable (y), and GPA, Age, ...
2
votes
0answers
48 views

Downsample a stochastic process without losing correlation statistics?

I have a stochastic variable $X(t)$ which changes at a discrete set of random times $t_1, t_2, \dots$. I can simulate this stochastic process to obtain a series $X(t_1), X(t_2),\dots$ However, the ...
1
vote
1answer
909 views

QR decomposition computational efficiency

I am struggling to find a reference for this: In terms of big Oh notation does anyone know of any expressions for the computational time taken by commonly used algorithms for QR decompositions?
-1
votes
2answers
2k views

K-Means clustering and correlation

I ran K-Means on my dataset, it's a small dataset of 200 countries x 6 export sectors. My results formed three clusters. Now I want to check whether these three clusters are correlated with another ...
4
votes
1answer
680 views

What's a mean field variational family?

I'm working through variational Bayesian methods at the moment, and I think I have a grasp of the bigger picture. Where I sometimes have trouble is with the exact details of how it can be implemented. ...
2
votes
0answers
41 views

Using Markov random field spatial weights to account for spatial autocorrelation

I am looking at the relationship between life expectancy and smoking rate within the London boroughs. I thus created a bayesx spatial regression model including a term which assigns spatial ...
0
votes
3answers
3k views

Proof for “The sum of the observed values $Y_i$ equals the sum of the estimated / fitted values $\hat Y_i$” [duplicate]

I needed some help trying to understand why the sum of the observed values $Y_i$ equals the sum of the estimated values $\hat{Y}_i$.
3
votes
1answer
463 views

How does scipy calculate the binomial CDF so fast?

I ran scipy.stats.binom.cdf(500006, 1000000, 0.5) and it took less than a milisecond. This is crazy as binomial CDF involves summing up a bunch of binomial coefficients. What approximation algorithm ...
1
vote
1answer
43 views

Is it methodologically correct to, in some specific situations, merge related binary variables into one continuous?

I have 10 yes/no binary variables regarding different types of economic risks associated with running a startup. I have a database of around 100 different startup companies evaluated on each of those ...
0
votes
1answer
33 views

Which analysis method question

I am currently conducting a short research paper for my university course and am struggling to determine which statistical analysis would be appropriate for the data I have collected. I surveyed two ...
2
votes
1answer
70 views

Covariance of Constrained Maximum Likelihood Estimators

I plan to numerically estimate the parameters of a GLM but with constraints imposed on some of the parameters. In this case, does the general approach of estimating the covariance matrix of my MLE ...
3
votes
1answer
113 views

Obtain global linear regression estimate from subsamples

I want to estimate $\widehat\beta$ in a simple linear regression with scikit. $$y = X \beta + \varepsilon$$ The problem is that the dimension of the complete $X$ is too large to fit into memory. Is ...
2
votes
0answers
31 views

Simulation - problem of maximization inside a circle

I am doing some projects related to statistics simulation using R based on "Introduction to Scientific Programming and Simulation Using R". In the Students projects session (chapter 24), I am doing ...
6
votes
2answers
907 views

Subtracting very small probabilities - How to compute? [duplicate]

This question is an extension of a related question about adding small probabilities. Suppose you have log-probabilities $\ell_1 \geqslant \ell_2$, where the corresponding probabilities $\exp(\ell_1)$...
0
votes
0answers
105 views

Must MSA be re-performed after annual instrument calibration?

I'm working in an automotive company that must follow IATF (International Automotive Task Force) requirements. I administrate a test-bench that performs radio frequency tests using an appropriate ...
2
votes
1answer
433 views

Vectorised computation of logsumexp

In this related post there is an explanation of how you can add together two very small probabilities using the logsumexp function, and how this can be programmed into base ...
2
votes
1answer
334 views

Data perturbation with normal variables

I am doing some projects related to statistics simulation using R based on "Introduction to Scientific Programming and Simulation Using R". In the Students projects session (chapter 24), I am doing ...
0
votes
1answer
77 views

Correcting data using poisson-regression

I'm new to stats and I was wondering if anyone had any good resources that could explain to me: How one can correct their data (false-positives) using Poisson-regression. I've been looking for some ...
3
votes
1answer
58 views

Computing Wassertein Distance

For two probability measures $\mu$ and $\nu$, the Wassertein Distance is defined as $$W_p (\mu , \nu) = \left[ \inf\limits_{\gamma \in \Gamma} |x-y|^p \, d\gamma (x,y) \right] ^{\frac{1}{p}} \, , $$ ...
0
votes
1answer
512 views

Propensity to pay modelling

I am trying to build a propensity to pay model given an intervention to a customer. Context: The population I am dealing with are customers who were supposed to pay some amount on a certain date ...
1
vote
1answer
96 views

Tracking Moving Objects with Kalman Filters— Over-fitting over time?

I've been learning about Kalman Filters, and the classic example given is tracking an object via radar/gps. My issue here is that each time you get a new data point, you update the error in the ...

1 2
3
4 5
13