Newest Questions
219,234 questions
0
votes
0
answers
4
views
Fraction selection framework for validating daily Biomass output in a Biophysical Model
I need to select the best fraction for my biomass model output through validation with field data. This is what I have:
I have a model that outputs daily Biomass values
The model output is expressed ...
0
votes
0
answers
7
views
Can DiD estimator be used for Association rather than Causality
If for example, we have this regression: Y = a + ß_0 + ß_1 Treatment + ß_2 Post + ß_3(PostTreatment) + e
Variables: Y: weekly growth rate of decease cases per 100k p., Treatment: 1 countries with face ...
2
votes
0
answers
26
views
When is likelihood inference metric estimation?
I recently learned that realizations of a stochastic model that is commonly used in population genetics, the Kingman coalescent, can be viewed as random metric spaces. In particular, each realization ...
1
vote
0
answers
11
views
How to model time series with uncertainty using probabilistic techniques?
I'm reaching out for guidance on probabilistic modeling for time series data. My background is not in machine learning, so I'm looking for insights from more experienced practitioners in the field.
My ...
3
votes
2
answers
33
views
How to calculate risk or prevalence difference using svyglm
I am trying to calculate risk (or prevalence) difference using svyglm. After searching for relevant documentation, I still feel uncertain if I can use a general ...
0
votes
0
answers
8
views
Is it possible to train/calibrate a conformal prediction model with a batch where some classes are missing?
I am working on a conformal prediction task with python.
I am currently using the crepes package but I encountered similar problems with mapie.
Consider a sequential learning task, where:
we define a ...
3
votes
1
answer
98
views
Is it possible to find individual estimates for phase and magnitude in this communication problem?
In a wireless communication scenario, received signal $y$ is governed by the following equation:
$$y=hx+n$$
where $h=|h|\exp(j\theta)$ and $j=\sqrt{-1}$ is the random channel coefficient with distinct ...
0
votes
0
answers
4
views
How to Measure Heterogeneous Effects in Conjoint Analysis
I am conducting a conjoint experiment, and my theory predicts that treatment effects vary by respondents’ socioeconomic status—specifically, whether they are poor and employed. I have approximately 1,...
0
votes
0
answers
18
views
Doing statistics with incomplete causal variables
I am broadly curious about making the following notion precise, and getting answers:
How can we max-min our confidence that we've identified causal factors
that should be part of a problem model, ...
4
votes
1
answer
59
views
Response derivatives from GAMs at different levels of a second predictor
I've been trying to work on some GAMs recently and look at the response derivatives at different factor levels. gratia::response_derivatives() seems like the way to ...
0
votes
0
answers
31
views
neophyte python coder needs hand held analyzing data [closed]
So, I have this scientific instrument, and I'm getting great data out of it, but as all real apparatus are prone to, the signal can have lots of noise mixed in.
To combat that error, I've implemented ...
0
votes
0
answers
38
views
PCA and lmer handling zeros
I am conducting an evolutionary study in which I am taking various measurements of spines from related species of dragonflies under different predations. Some species do not have all spines, others ...
3
votes
1
answer
143
views
Error metrics for time series models when data has a small range
My time series takes values in the range [-3, 3]. Standard percentage error metrics like MAPE don't work well here.
For example, if the actual value is 0.1 and I predict 0.3, that's a 200% error even ...
1
vote
0
answers
58
views
How to determine if your analysis in causal, somewhat causal or non-causal?
I'm analyzing monthly time series data to estimate the effect of a policy intervention.
I have approximately 5 years of monthly data (60 observations), with the intervention occurring roughly in the ...
0
votes
0
answers
29
views
Intuition for using "Chaotic" spatial weights in covariance estimation (vs Uniform or Gaussian)?
I am working on feature extraction for texture analysis. I am trying to understand the physical intuition behind using Deterministic Chaotic Weights compared to standard weighting schemes.
1. Uniform ...