Stack Exchange Network

Stack Exchange network consists of 175 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

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.

0
votes
0answers
9 views

Modeling the mean with a more complicated model than a simple average

Marc Kery discusses modeling means as an alternative/can be synonymous to a simple average in his book: Kéry, M. (2010). Introduction to WinBUGS for ecologists: Bayesian approach to regression, ...
0
votes
0answers
8 views

Is this variable suitable for a categorical regression (multinomial logistic regression)?

I have created a dataset starting from a series of multiple choice (3 choices) questions. ...
2
votes
0answers
19 views

Accuracy of Keras Model is Very Low for Identifying Differently Colored Objects

I am using transfer learning approach to train my keras model to identify objects which have same structure but the colors are different i.e objects are to be identified by their respective color. ...
0
votes
0answers
6 views

Calibration plot when majority predicted probabilities are <0.1

Let's say I built a prediction tool out of 10 000 cases with a range of baseline predictors. The end point occurred in 5% of cases. Great majority of predicted probabilities is below 0.1 and only ...
0
votes
0answers
12 views

What to do when variance increases with sample size

I have a large dataset that has been compiled from 6 surveys. The values are how much of a food additive is used in a particular food. After breaking the additives into pre-established groups, I am ...
1
vote
1answer
42 views
+50

How to calculate conditional hazard at time zero

I am looking for a statistical method to calculate the conditional hazard before the diagnoses or study has started; in other words, time is not a variable. For example, here is a hypothetical ...
7
votes
3answers
178 views

How to reduce predictors the right way for a logistic regression model

So I have been reading some books (or parts of them) on modeling (F. Harrell's "Regression Modeling Strategies" among others), since my current situation right now is that I need to do a logistic ...
1
vote
1answer
17 views

Inverted dose-response variables

Context: Often when we carry out dose-response modelling we want to estimate the dose required to elicit a predetermined response (i.e. response ~ dose). Typically this is done with inverse ...
0
votes
0answers
11 views

What's the purpose of `tunegrid` in the caret package?

I am deciding the parameters for a random forest classification model. I am using the caret package and read, here, about the tuning grid. My understanding is that it helps determine the best values ...
0
votes
1answer
22 views

improving accuracy of classification model

I have data with 95 numeric variables and 5 categorical variables. My Y has 2 values. I built a decision tree and my accuracy was 81.8%. Then I created 3 new variables as follows. It improved accuracy ...
0
votes
0answers
11 views

Updating training data with predicted data

Is it valid to update a model with data that has been predicted? For example I train a model on 10000 records. Later I obtain new data requiring prediction, say 500 records, I use the model trained ...
0
votes
0answers
17 views

Predicting whether a potential business agreements will be won or lost [on hold]

I am currently working on a project using a sales system and trying to come up with a way to use the current pipeline of potential sales to predict if the business agreements will win or lost based on ...
0
votes
0answers
19 views

High Accuracy- Seems fishy [closed]

I am trying to build a Supervised Classification based Predictive Model. The data consists of 13 qualitative variables. I built a predictor based on three columns and now I am trying to apply Logistic ...
0
votes
0answers
10 views

Categorical variable postestimation at cluster level

I have a question about categorical variable regression and post-estimation procedures. My aim is to estimate the “probability of success” (say, odds-ratio) at an aggregate level using a lower-level ...
0
votes
1answer
32 views

Defining the cases where Neural Networks outperform tree-based methods

It is well-known that neural networks are currently superior to most of the alternatives to do prediction from images (with CNNs) and sequential data (RNN, transformers...). However for other ...
-2
votes
1answer
19 views

Revenue Projection

Giving that we have Monthly revenue data for pass 3 years (36 rows of revenue) We have other data including economic indicators, industry indicators as well (other columns in the 36 rows) ...
0
votes
1answer
66 views

How can I better predict with (g)lmer with missing values?

Suppose I'm building a mixed model in R, and I want to use that model to predict new data for which I might not know the value of all the features. Or in some cases, it might not be so much that I don'...
0
votes
0answers
12 views

How does multicollinearity affect the feature selection process?

I have a classification problem with a modest number of records (approx. 10,000) and dimensions (30 dimensions, 25 are categoric and 5 are numeric). The response variable has two classes (T/F). I'm ...
0
votes
0answers
9 views

Binary target prediction using LSTM with sparse events in time

I have a data of patients that have multiple events happening in there medical history, I'd like to predict a target of having a specific targeted-event in the next 30 days. The data is timestamped ...
2
votes
1answer
21 views

Why are AUC and logloss metrics not available in the “maximum metrics” table produced by H2O? [closed]

I am running the h2o.gbm algorithm using five-fold cross validation to predict a binary outcome. I want to see what threshold to use as a cutoff for classifying predictions, and I am wondering why the ...
0
votes
0answers
8 views

Predictive value of a predictor - model choice

I have been stuck on this for a while. I am supposed to find the "predictive value" of a predictor - entrance exam score. Together with the entrance exam score also grades from the first semester are ...
0
votes
0answers
14 views

How do I diagnose collinearity with rfe() from the caret package?

I have 12,000 records and I"d like to predict a two-class outcome. I'm deciding which predictors to keep and I'm having trouble with two problems. 1- I get an error message because I have categories ...
0
votes
0answers
8 views

Do I need to do glmnet after doing a cv.glmnet?

I'm studying now about the model selection from the ISLR book. I'm don't understand about whether should I do glmnet() after I do ...
1
vote
1answer
8 views

Penalizing Some Errors more Than Others

Suppose you want to penalize a model when it makes mistakes in classifying some points in a test set more than other points in the test set. How would you do this? Would you just use another measure ...
0
votes
0answers
28 views

interpretation of ordered logit regression with categorical independent variate

I would like to predict the quality of plants in certain area. I divided the quality of the plants in 5 groups; 0 to 5. And we've measured 5 different areas; control, 1, 2, 3, and 4. I ran a logit ...
0
votes
0answers
18 views

Use R2 in hypothesis testing

I am conducting a large simulation study, where different statistical models are set up from training data and cross validated on a validation set. I want to assess, under which conditions which ...
1
vote
1answer
28 views

Logistic regression - handling very few observations for a level of a categorical variable

Context I'm not sure what to do in the scenario where one of the levels for a variable has so few levels that there's a good chance splitting the data into $70\%$ training $30\%$ testing will result ...
0
votes
0answers
14 views

Splitting data into test and training when there are a low number of observations for a level within a variable

Context If a level within a categorical variable has less observations than there are elements in the training set then there's a chance that all of the elements of that level will be contained ...
0
votes
0answers
6 views

Different values in test / training data variable

The code and error in the following MWE represent my issue. code ...
1
vote
0answers
18 views

80-20 better than full dataset for LightGBM

Recently I have been using LightGBM as regressor in order to predict, on a dataset of 20 thousand observations. I have two modes, 1) Production and 2) Testing. The first one just trains a model with ...
0
votes
0answers
12 views

Predict hazard relative to a specific sample

I want to predict the hazard ratio and 95% confidence intervals relative to a specific sample rather than the population mean: ...
1
vote
2answers
40 views

General question: How do you visualize/deal with a lot of predictors?

In STATS classes, one is always taught to draw a picture to look for outliers, to look for the distribution type, to look for patterns in general. However, when you have a dataset with a lot of ...
0
votes
1answer
25 views

How to give more importance to one variable in a logistic regresion model? [closed]

I'm adjusting a logistic regression model for prediction, but if the person interested says: All variables are important for me, but especially X2 is more important. How I give that variable more ...
0
votes
1answer
28 views

Poor performance from `gstat::krige` with a noise predictor

I'm new to kriging, and I'm considering replacing a use of inverse-distance weighting (IDW) in a spatial modeling project (implemented with gstat::idw in R) with a ...
0
votes
0answers
10 views

Using posterior variable in credit risk model

I am rebuilding a credit risk model using logistic regression (either ridge penalty or elasticnet) to predict first payment default. Historically, the company approves an applicant for a loan to ...
0
votes
0answers
8 views

Prediction of dementia based on diagnoses

I have got this dataset consisting of records for primary care patients: Date of visit (btw 2013-2017) ID Diagnosis (ICD10) registered at the visit A flag indicating if a patient was diagnosed with ...
1
vote
0answers
48 views

Longitudinal Data with Equal Outcomes Within Individual Samples

I need to prepare some data for plugging into a predictive model. The data is in tidy format, but it comes from an audit table, i.e. every change made to a record is recorded and stored as a separate ...
-1
votes
1answer
30 views

How to predict without labeled data

I have a data set for patients visiting emergency departments containing following features: The output variable in this data set is "disposition" - whether a patient becomes admitted or discharged. ...
0
votes
0answers
13 views

Choosing prediction model with regularization, spatial cross-validation and bounded predictions

I am new to machine learning and R. I want to run a statistical model to predict daily hours of supply of electricity (y). I have several x variables to use for prediction. I have three goals to ...
0
votes
1answer
31 views

Using outlier records as a feature in model building

I am exploring the Big Mart Sales III dataset and trying to understand if using outlier rows to build a feature for predictive modeling is a sound and correct approach. This is how I have proceeded ...
0
votes
0answers
15 views

Model evaluation - High sensibility and specificity but low MCC

I trained a Random Forest classification model to predict bioactivity for different protein targets. Both my training and test sets were highly imbalanced with ~99% of the majority class. Now that I'...
0
votes
0answers
11 views

There exists a k-means like method that allows good predictions when we have two sets of variables (one dependent and one independent)?

Suppose you have a set $X$ of dependent variables and a set $Y$ of dependent ones observed on $N$ individuals. So, I have a vague idea that a causal relationship should be validated(or measured, or ...
1
vote
0answers
74 views

Why not use (nested) cross-validation to update weights when building final model?

I have been trying to find an answer to this question for some time. I understand that cross-validation is primarily used for model selection, i.e. to tune parameters/hyperparameters, but I don’t ...
0
votes
0answers
37 views

How to interpret the Matthews Correlation Coefficient (MCC) for an imbalanced data set

I am trying to assess the performance of a machine learning model that has been passed down. The XGBoost model was trained on data that had a class imbalance of 84% majority class (label 0: 117,409 ...
0
votes
0answers
9 views

How to create a model that will display a graph/table showing predicted units sold?

Goal: Create a model that will display a graph/table showing predicted units sold and then be able to use that to compare actual units sold during the season, eventually turning into an accumulated ...
1
vote
1answer
42 views

How does facebook prophet handels missing data?

Prophet's paper (forecasting at scale by SJ Taylor - 2017) says the following on missing data: "Unlike ARIMA models, the measurments do not need to be regularly spaced, and we do not need to ...
1
vote
1answer
21 views

Predicting items based on time and weather

I've got a data-set with items bought, the time it was bought (I can add the weather of the location at that time of the day). I would like a simple "prediction" model based on time and weather. ...
3
votes
1answer
35 views

Solve the optimization problem of tree, should we make each rectangle contains exactly one training data point?

I was reading the book "An Introduction to Statistical Learning with Applications in R". In page 306, when talking about the objective function of tree model, the book says: "The goal is to find ...
0
votes
1answer
15 views

How to prove High Sampling Variance in over-fitted functions

I've been reading recently about over-fitting and it is frequently related to High Sampling Variance and Low Bias characteristics. However, what is the metric used to state the High Sampling ...
1
vote
1answer
34 views

Using GDP to predict Probability of Default [closed]

I would appreciate if you could help me to answer on how to use the GDP (Gross Domestic Product) to predict the probability of default (Probability of corporate defaulting in their payment). By using ...