Newest Questions
218,408 questions
9
votes
2
answers
713
views
Preservation of mean>median for sums
Suppose $X$ and $Y$ are non-negative real random variables and that the mean is strictly greater than the median for each of the two variables. Does it follow that the mean of $X+Y$ is strictly ...
- 51k
1
vote
1
answer
43
views
What would be the distribution of days with n messages in large group chat?
I have the data of a large chat group (~100 participants) and relative long time (since Oct. 2017 until April 2025). I have made the histogram for the number of daily messages and I am a little ...
- 111
0
votes
0
answers
27
views
Cramér-Rao Lower Bounds for angles and other bounded estimators
The Cramér-Rao lower bound gives a lower bound on the variance of unbiased estimators. However, I have only ever seen it used on variables that can take on any value in $\mathbb{R}$. Is there a ...
- 1
0
votes
0
answers
16
views
Is this a sound way to apply Bayes’ rule to track confidence in process improvement over time based on observed progress (given the assumptions)?
Problem
We're trying to improve our defect detection system in a manufacturing process over the course of a year. The goal is to gradually increase the percent of defects caught earlier in the process....
- 301
5
votes
2
answers
215
views
Finding UMVUE of a Poisson parameter
Suppose $X \sim \operatorname{Poisson}(\lambda), \lambda>0$ . Let $\theta=\lambda^3e^{-2\lambda}.$
a) Show that $S=(-1)^{X-1}X(X-1)(X-2)$ is an unbiased estimator of $\theta$.
b) For a random ...
- 321
0
votes
0
answers
19
views
Minimum number of periods to make use of time fixed effects
In econometrics entity and time fixed effects models are often used to study longitudinal panel data. One of the most interesting uses of this is that the time fixed effects can isolate the unobserved ...
- 223
2
votes
0
answers
22
views
Using z-values to determine significance of linear regression with non-normal residuals
I have a linear regression with non-normal residuals for which I used robust standard errors. I was told to use z-values to determine coefficient significance but I'm struggling to find any resources ...
- 21
2
votes
1
answer
22
views
Best test for multi-dimensional data with variable number of members
I have a pretty large dataset centered around 1000 conversations and I am struggling to find the best test to run to find statistical significance in the difference between my results, as any lessons ...
- 21
1
vote
0
answers
11
views
How to train a embedding model so that I can customize how it'll be stored in vector space?
I'm trying to create custom embedding model that can embed robotic actions into vector space. Traditional embedding model only have an option to convert text to vector. Robotic actions are much more ...
- 11
0
votes
0
answers
21
views
Neural Netowrk as an Agent doesnt learn [closed]
The code below has a bug somewhere, but I don't know where, so I would like someone to take a look. The agent never learns — it just lands between -200 and -400 reward. I’m not sure if I’m mishandling ...
- 1
1
vote
1
answer
59
views
Custom contrasts involving interaction terms using emmeans
I'm trying to setup a custom contrast using emmeans but am a bit unsure on how to do so properly.
I have two factors, let's call them A and B, with three levels each.
I want to test the following ...
- 11
7
votes
2
answers
234
views
Expectation of a random variable
In a game, a player starts with an initial score $0$. The player rolls two fair dice, and the difference between two outcomes (say $X_1 and X_2$) is noted. If the difference $D = |X_1-X_2|$ is ...
- 321
0
votes
0
answers
74
views
ARIMA Modeling on Aggregated Global Data: Stationarity, Differencing, and Forecasting Concerns
Good day! I’m currently working on a time series analysis using ARIMA for a global yearly dataset. The data comes from a publicly available and anonymized dataset (obtained from the Global Burden of ...
- 31
1
vote
0
answers
22
views
input oriented Bootstrapping data envelopment analysis bias correction efficiency greater than one
In the input-oriented DEA model, efficiency scores typically range between 0 and 1. However, when using bootstrapping and bias correction for the estimator, some units have efficiency scores that ...
- 11
1
vote
3
answers
130
views
Can a biased hypothesis test be preferred over an unbiased one?
In point estimation, a biased estimator can be preferred over an unbiased one considering bias-variance trade-off. Is this also the case for hypothesis testing? More generally, can a biased hypothesis ...
- 109