Questions tagged [predictive-models]
Predictive models are statistical models whose primary purpose is to predict other observations of a system optimally, as opposed to models whose purpose is to test a particular hypothesis or explain a phenomenon mechanistically. As such, predictive models place less emphasis on interpretability and more emphasis on performance.
13 questions from the last 30 days
1
vote
1
answer
100
views
How to compute confidence and uncertainity of model without ground truth from softmax output?
Suppose I have 3 classes A,B,C.
Performing:
y_pred = model.predict(X) # suppose X only two sampel
Returning vector with length ...
0
votes
0
answers
15
views
Comparing probabilities of two models
Consider a dataset and two binary classes CLASS_A and CLASS_B, with different proportions of 0 and 1. Suppose we train a model such as XGBClassifier for both ...
- 1,690
1
vote
0
answers
9
views
Risk score developement from admin health data: when the test for the risk factor and outcomes are confounded by indication
I have health records of immunodepressed patients who may have event histories like
[high risk demographics] -> [low lymfocyte count] -> [high viral load] -> [clinical events]
From those data ...
0
votes
0
answers
7
views
Prediction with categorical data in semopy [migrated]
I'm using semopy for the first time (I am more familiar with lavaan in R). I was able to apply the ...
- 119
1
vote
0
answers
26
views
XGboost classification order of features influencing predictions [closed]
I am creating a UFC fight prediction model using xgboost classification, trained on a dataset that includes the stats of 2 fighters and the result of the fight. After training a model, I defined a ...
- 11
0
votes
0
answers
39
views
Select best predictor from multiple ROC Curves in R
Packages
library(ggplot2)
library(dplyr)
library(caret)
library(plotROC)
library(pROC)
library(ROCR)
3 regression models
...
- 1
1
vote
1
answer
62
views
Taking the sum of predicted probabilities from logit model?
I am using a logit model to predict the probability that students pass a particular course. I run the logit, generate predicted probabilities for the students in my sample, and want to compare the ...
0
votes
0
answers
15
views
Rubin's rules Predicted values from mice imputed datasets and rms library
I am starting an analysis midway so my starting point is multiple imputed datasets created using the mice.
...
1
vote
1
answer
29
views
Optimal threshold for a predictive covariate on treatment effect
here a physician with a moderate understanding of statistics and am seeking guidance on analyzing a continuous covariate (biomarker) that I suspect is predictive of treatment effect. My analysis ...
- 31
8
votes
4
answers
554
views
Signal-to-noise ratio in predictive modeling and machine learning
The interesting comments to this question get into how signal-to-noise ratio plays into ability to make predictions. Being more explicit about it, how does signal-to-noise ratio factor into how good ...
- 67k
2
votes
1
answer
158
views
predict() function fails for lmer in R when NAs present in dataset
The issue is not how to format/obtain data but how to run predictions for linear mixed effect model for given set os estimated fixed effects in case of NAs present in the data.
The predict() function ...
- 316
0
votes
0
answers
18
views
Substitution modeling under simulated inventory constraints
In the context of retail, substitution is typically defined as "when customers that prefer product A, but when unavailable, purchase product B, instead."
Naively, the most of useful ...
- 3,520
0
votes
0
answers
26
views
Simulating bias comparing linear regression with Cox Proportional Hazard model for prediction on censored data
I wanted to show how censoring biases linear regression and how that could be resolved with survival regression using simulated data. My simulation now suggests, that survival regression (I tried a ...
- 281