All Questions

0
votes
0answers
6 views

Interpreting LM Model Ouput

I know interpretation of model output and selection has been discussed a bunch, but I wanted to ask / better understand the question below more directly. Below is the output of a model selected using ...
0
votes
0answers
3 views

Dealing with superexponentially raising losses when training a neural network

Sometimes losses contain functions which raise very fast - exponentially or even superexponentially. This can cause overflows to +∞. What is the best method of dealing with such bad-behaving losses?
0
votes
0answers
5 views

Solving the Binary Logistic Regression with LASSO penalty

The objective function of the Binary logistic regression with the LASSO penalty is given by, $argmin_{\beta_0,\beta}$ { $-{1}/{n}$ $\sum_{i=1}^n (Y_i({\beta_0}+{\beta^T}x_i)-log(1+exp({\beta_0}+{\...
0
votes
0answers
5 views

Should I convert the number after log transformation back to the original for calculating RMSE?

When I built the model, I applied the log transformation to all variables including the dependent variables. Now, I'm calculating the RMSE for the evaluation, and the result is in the log format, ...
0
votes
0answers
7 views

K-means gives non-spherical clusters

I am trying to cluster 24 month utilization behaviors of customers using sklearn/K-means in python. When I plot the customers by clusters in a 2-D space (Principal Components 1 and 2 of my 24-point ...
0
votes
0answers
4 views

Party package (ctree/cforest) in R with NA values

I don't quite understand how ctree/cforest deal with missing predictors. Can someone please explain this further?
0
votes
0answers
7 views

Peacock Test Implementation in R [on hold]

I have two empirical distribution (without access to the observations themselves). So, in order to see if the empirical distributions are "significantly" different, I took two large random samples ...
0
votes
0answers
6 views

Matrix transpose lemma [on hold]

I have a matrices $A$ and $B$ which are both $(n \times p)$, a vector $Y$ which is $(n \times 1)$ and a vector $Z$ which is $(p \times 1)$. I know that $A'A=B'B$. I want to show that: $$||Y-AZ||^2_2=...
3
votes
2answers
13 views

Understanding Lasso Regression's sparsity geometrically

Whenever someone writes about Lasso and Ridge Regression thy draw this diagram with the circle or with the diamond. In the case of the diamond (Lasso regression) it is then always stated that Lasso ...
-1
votes
0answers
21 views

How to explain the result of this polynomial equation?

Lets say that I predicted plant productivity (logy) by precipitation (logx1) and moisture content of soil (logx2). my original data gives me the best result only after taking logarithm on both side of ...
0
votes
0answers
4 views

How to calculate weights for a weighted ordinal regression, with > 2 possible outcomes (unbalanced)?

I'm trying to perform an ordered logistic regression with the response variable being a certain 'level' of achievement (level 1 = highest level, level 2 is one below, until level 10 - the lowest level)...
0
votes
0answers
3 views

Sampling bias in multinomial logistic regression

I am interested in estimating a set of coefficients in a multinomial logistic model. However, I only observe a subsample of the true sample in which base category $A$ was chosen. I have no way of ...
-1
votes
0answers
7 views
0
votes
0answers
6 views

Reduce scoring time for test sample of K-Nearest Neighbors regression for a time series dependent variables

My dependent variable is values of $ balance amount for an entity over next 24 months after the entity is subjected to a specific treatment. I am trying to predict these 24 values (bal1-bal24) for ...
1
vote
1answer
11 views

why does random forest trees need to be deeper than gradient boosting trees

in Elements of Statistical Learning chapter 15. Random Forest, we see authors' note on RF v.s. GBT. One of them is that at 1000 terms, GBM depth 4 has smaller error than RF depth 6. Also we notice RF ...
0
votes
0answers
19 views

If T is a sufficient statistic for theta, is H(theta|x) = H(theta|T(x))?

I was trying to prove that sufficient statistics attain equality in the data processing inequality by a slightly different route than I usually see, and came across an odd expression. (I care more ...
0
votes
0answers
14 views

Train, validate and test on different folds

To elaborate on the title. Does it make sense to do the following: Split the data into train and test set. Cross-validate on the training set. Remove redundant variables, tune parameters of the model ...
-1
votes
0answers
23 views

How can i fit the following model in R ? What am i missing?

Picture shows the regression from Daniel and Moskowitz (Momentum Crashes) Rwml is the monthly log return So the first column is clear, i got nearly the same values, at least the same magnitude. But:...
0
votes
0answers
17 views

What does “density” really mean in probability density function in statistics? [duplicate]

I am familiar with the concept, but I simply can’t get my head over the intuition behind it. While being a derivative, it describes the rate of change for one unit. Simply put, we can say that it ...
0
votes
0answers
14 views

Which function can approximated with Neural Networks using only linear activation functions?

I want to find out which functions can be approximated up to arbitrary accuracy using Neural Networks with only linear activations. On this page I found out that with linear activation functions, the ...
0
votes
0answers
10 views

Difference-in-differences analysis with multiple time periods

I am at the effects of a law on listed companies. I have a control group and a treatment group, but instead of having a pre treatment period and a treatment period I have 3 time periods. I have ...
0
votes
0answers
10 views

Should I check structural breaks when forecasting, given limited time window

I attached the graph of a time series below, which is a series of the probability of defaults. This is a quarterly time series. When I am using a regression model to forecast this time series, should ...
1
vote
1answer
12 views

Logit model with all significant coefficients, but low prediction capabilities

I'm facing this issue with a logit model with R. I got my coefficients significant (already removed the Linear dependent variabiles and the not significant variables), as: ...
-1
votes
0answers
14 views

Too many categorical variables

I have been working a data set that contains some features of used cars and their prices. Features like number of cylinders, millage, model, brand and type. Brand and type have a strong impact on the ...
0
votes
0answers
11 views

Is there some mathematical logic behind any differences between stock and delta? [on hold]

I am working on a project where I am validating carbon stocks at two different points in time. I am using allometric equations/growth and yield model to predict carbon stocks in forests. The model is ...
0
votes
0answers
11 views

The relative size of the SE for ratios compared to the SE for means

Pardon a relative novice's question. I'm seeking a reference that describes, compares, and gives formulas for the standard error for ratios and the standard error for differences between means when ...
0
votes
0answers
5 views

factor analysis of mixed data - difference between quality and contribution

I started to use the R package FactoMineR to perform factor analysis of mixed data (function FAMD). Could someone please be so kind and explain the difference between quality (cos2) and contribution (...
2
votes
0answers
25 views

Do probabilities exist in reality? [on hold]

I came across the seemingly interesting work of William Briggs in relation to probability and statistics. He make some bold claims, like the following: Probability does not exist as a thing, as a ...
0
votes
0answers
5 views

Using mutual information for feature selection between feature maps (python)

I want to do feature selection between 512 3X3 feature maps from convolutional layers. I want to calculate a 512X512 MI matrix and choose 256 feature maps with the lowest MI values between themselves (...
1
vote
0answers
14 views

determining confidence interval for point estimate from regression

Given a regression equation and confidence intervals for the coefficients, how do you compute a confidence interval around a point estimate derived from the regression equation? I'm specifically ...
0
votes
0answers
4 views

where to specify covariates in a linear mixed effect model

I am using lme4 to create a mixed model for my data. I am looking at the effect of land cover on soil properties at three depths, for example, carbon concentration. I have identified my fixed ...
0
votes
0answers
8 views

Interaction not significant in Cox regression model?

I ran a cox model with an interaction between two categorical variables (sex and treatment group) included like: Coxph(Surv(time, status) ~ sex*treat+age+edu, data=data) And there is no significant ...
0
votes
0answers
14 views

How to interpret quadratic effects?

I have built a LMM (using LME4) for understanding how investment fund's structural characteristics (things like how much debt they have, their size etc) impacts upon performance. In the analysis, I ...
0
votes
0answers
20 views

Normal versus student-t distribution

I estimated two GARCH models, one with the Normal distribution and one with the Student-t distribution. The conditional volatility shows less noise for the model based on the Student-t distribution. ...
0
votes
0answers
10 views

What are the ways to generate multivariate data? [on hold]

I am trying to teach myself statistics and I was curious how to generate multivariate data. Say, given 1000 multivariate observations I want to generate 10,000 observations. If I randomly sample these ...
0
votes
0answers
9 views

How to get a threshold value?

I have a dataset of genes and their frequency in the dataset as: Target_Gene:A,B,C,D,E,F,G,H....,Z Count:260,150,120,90,50,40,36,27,.....,1 Now I take a Sample from this: Target_Gene:A,C,F,J,K.......
0
votes
0answers
5 views

(plm in R) Regression Model Using Column Numbers or Generally Avoiding Manually Entering Variables? [on hold]

I'm learning R as I work through a project and I believe that there is a simpler method to do what I want than to enter every single value in the model. I am using the plm function for fixed effects ...
0
votes
0answers
9 views

TF-IDF versus TF for Cosine Similarity

My task is to examine emails and determine the cosine similarity between pairs of emails to see which ones are the same or almost the same. I was thinking of using the TF-IDF technique, but am not ...
0
votes
0answers
8 views

Generating y label in an sklearn pipeline?

I have a simple use case - I want to, as part of an sklearn pipeline, generate y labels based on X features (e.g. for predicting some signal in X, n timesteps in the future). I want it to be a ...
1
vote
0answers
4 views

Clarify Partial Correlation and GLM

This has been plaguing me for a while. Let's say I am exploring the association with fasting glucose and BMI and am returned with a significant positive correlation. Now, I want to adjust for age and ...
0
votes
0answers
7 views

General forms of conditional heteroskedasticity

I am currently reading a paper published in the Journal of Finance: Rapach et al. (2013), 'International Stock Return Predictability. What is the role of the United States'. The authors quote they ...
0
votes
0answers
3 views

Do we have to make cycles and trend covariates/predictors stationary in order to use them as valid predictors in a Dynamic Regression Model?

I want to extract patterns from macroeconomic indicators for use in predictor a target variable. In particular, I plan to decompose the macroeconomic variables into trend, cycle, maybe seasonality and ...
2
votes
0answers
12 views

Which distance metric to use to cluster categorical sequences (clickstreams or clickpaths)?

For my research, I want to cluster website visitors based on their clickstreams to understand different information behavior patterns (i.e., customer/visitor journeys). The data can be characterized ...
0
votes
0answers
15 views

Is this an example of a T-statistic?

Excerpt from a study https://www.researchgate.net/publication/328838624_No_More_FOMO_Limiting_Social_Media_Decreases_Loneliness_and_Depression: Although not every subject complied perfectly with ...
0
votes
0answers
5 views

What is the most appropriate model to detects fault for UAV's sensor data with Autoencoders?

I want to use autoencoders to detect faults in UAV sensor data. So far, I found out that LSTM can be the most appropriate autoencoder model for this purpose. Is it so? Or, what can be the other ...
1
vote
0answers
8 views

SVD versus RSVD

In the so-called incremental SVD used for collaborative filtering: http://www.machinelearning.org/proceedings/icml2007/papers/407.pdf http://www2.research.att.com/~volinsky/papers/ieeecomputer.pdf ...
0
votes
0answers
13 views

Applicability of Markov Models for predicting user input

I am trying to predict user action based on the shown content of different modules. Lets assume the user sees a page where several content may be shown or not. The next action of the user depends on ...
0
votes
0answers
4 views

Questions about calculate the first k-th principal components of VR-PCA?

Variance-reduced PCA VR-PCA, focus on the first component calculation, what about the k-th PCs? Any idea? Thank you!
0
votes
1answer
20 views

Is it ok to use polynomial model to fit logarithmically transformed data?

Can we use polynomial regression using logarithmically transformed data. I transformed both y and x1 and x2 variables by natural log. After this, I used polynomial regression to fit. when using poly(...
1
vote
1answer
16 views

Why no “extra coefficient K” in the formula of Laplace's law of succession?

Below is an example about the calculation of Laplace's law of succession: Suppose we observe $y$ responses out of $n$ binomial trials. Assuming the trials are indenpendent, with unknown response ...

15 30 50 per page