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.
4,621
questions
1
vote
0
answers
25
views
How can I condition an autoregressive vector model given past observations?
I'm trying to predict the scores of a soccer team in the next match using a VAR (Vector Autoregression Model). So my first attempt was to define the model as it follows:
$$g_{1,t}=\alpha_{1}+\beta_{11}...
- 111
0
votes
0
answers
7
views
Maximizing Number of Forecasts, but not too great a cost
I wanted to ask is there a name for the following problem:
Suppose, I have a prediction model, which makes a prediction, when certain preconditions are met.
Naturally, the goal is to maximize the ...
- 1
0
votes
0
answers
66
views
Predictive Analysis Modeling Using Multiple Datasets
I am working on doing some predictive analysis, and we are attempting to implement a schema that we have found in a white paper. I have been searching for a while, but have not been able to come up ...
- 1
2
votes
1
answer
44
views
How to use GAMs to test whether a prediction model is properly calibrated in R
Suppose we wish to test whether there is an identity relationship between some observed values and the predicted values from some model, and allow for non-linear deviations from this hypothesis. This ...
- 1,213
0
votes
0
answers
20
views
Customer propensity model : creating features from geographical data
I am creating a customer propensity model to predict whether customer will buy in the next quarter.
I have used transaction data for the last seven years to make predictions.
In the data I have state,...
- 213
0
votes
0
answers
17
views
Managing Time-Series Forecasting Prediction Length vs. Prediction Frequency during training
Hello Stack Exchangers!
I am working on a time-series problem that uses a neural net to forecast the price of electricity. What is needed is for the model to predict the electricity price 3 hours out ...
- 1
1
vote
0
answers
47
views
Combining N number of weighted probabilities
I have a question that I thought was simple but I can’t seem to find an answer.
Let’s say I have N number of models that will predict a binary event and i have an accuracy assigned to each model, for ...
- 11
1
vote
0
answers
31
views
Decision tree and relationship between test and train results over time
My goal is to predict the number of policies in 5 years. Therefore, I train a decision tree that predicts the termination probability for each contract in a given year.
Then, I use a test data set and ...
0
votes
0
answers
66
views
Does a distribution of predictions similar in the test dataset and in an unlabellised dataset means that the model will generalize well?
I have trained a logistic regression model on a train dataset. The prediction on the test set is very good, with a nice distribution of the predictions : a high peak around 0 for the true 0s and a ...
0
votes
0
answers
20
views
How to model EU Carbon Permits (EUR)?
I'm trying to model EU Carbon Permits (EUR) using monthly data($P_t$) as shown below.
My objective is to predict EU Carbon Permits price from now to next 20 years (240 months) using monthly data. i.e. ...
- 1
9
votes
4
answers
2k
views
Is it required to train the model in entire data after cross validation?
I have a model trained as follows.
...
- 213
0
votes
0
answers
13
views
How can I compare a percentage chance of a bad outcome to a binary result of a good vs bad outcome. Medical Statistics
Sorry if the question seems vague but this is my first post. I am currently comparing a unit's performance in managing patients. I have their actual outcome dichotomised into good or poor and I have ...
- 1
0
votes
1
answer
27
views
Evaluating/combining PDFs over time to predict future value
I am trying to predict a value over time. I have historical data that I have used to calculate PDFs for the change over various time intervals.
If I'm trying to predict the value at time T0 and start ...
- 373
0
votes
0
answers
6
views
Ensemble modeling strategy
Since the ensembling model requires the individual models to be different for effectiveness, can I run two xgb models with one model metric as recall and the other one's metric as precision and ...
- 213
0
votes
0
answers
24
views
Evaluating country-level indicators through ROC analysis
Good afternoon.
I have defined a binary classification problem. Classes are based on country-level data on child undernourishment whereby a positive class is assigned to a country whenever it ...
- 1
1
vote
1
answer
88
views
Design and contrast matrices for analyzing the effect of two factors of interest on protein expression, controlling for factors and covariates
I've posted this question on the Bioconductor support site but haven't received any responses so I figured I'd try here. The post was quite detailed so I'm copying it over mainly as-written. Please ...
- 25
1
vote
1
answer
61
views
Predictive modelling strategy for spatial interpolation: complex data structure, small sample size, but n > p
I apologise in advance for the long post, my questions are deeply interconnected so it felt wrong posting them as separate threads. Please note also that I edited this post to account for Florian ...
- 531
0
votes
0
answers
37
views
What is the conclusion from cross validation scores?
I am training a model, and I'm using an xgboost model.
I used the following piece of code to find the cross-validation score
score=cross_val_score(final_classifier, X, y, cv=5, scoring="f1")
...
- 213
1
vote
1
answer
176
views
What is the Formula for Prediction Interval in Multivariate Case? [duplicate]
I am using linear model to do prediction, and I would like to calculate my prediction's prediction interval, which, when there is only one predictor, is
However, my model has three predictors. What ...
- 123
0
votes
0
answers
122
views
Predict from Bayesian Network in R
I have a dataset whose variables are NUM type. I created two subsets train and test and a bayesian network using HC algorithm from test dataset. I want to predict a variable of the network based on ...
- 1
0
votes
0
answers
37
views
multivariate vs multiple time series - term explanation and forecasting model
Here is the time series I have: (p1s is an abbreviation of product 1 sales in dollars)
...
2
votes
0
answers
238
views
Different prediction plot from visreg versus ggplot(Predict())
The problem:
I am hoping to visualise interaction effects between a continuous and a dichotomous variable from a cox proportional hazard regression. When graphing this with different packages (visreg ...
- 21
1
vote
1
answer
42
views
subscription prediction model- how to define users that didn't have an active subscriber
when I have created the training dataset for the subscription classifier, a user with an active subscription was labeled with 1.
I have a dilemma for users with label zero. At first, all the users ...
- 301
0
votes
0
answers
22
views
Handling new customers in customer propensity model
I'm using last four years' data to predict whether they will buy or not buy in the next quarter.
One problem I'm facing is customers who are not four years old. Is it right to keep them in the data ...
- 213
1
vote
0
answers
42
views
What is the best way (in terms of performance and accuracy) to perform value estimation/prediction?
How do I predict the value of an item given its features and attributes? What is the best approach to this regression problem, in terms of performance as well as accuracy? And how do I process this ...
- 111
2
votes
1
answer
62
views
Survival analysis when event cannot happen after timepoint
People are contacted for a survey for a fixed number of days. For simplicity, let's say this happens over 5 days. People can complete their survey on day 1, day 2, ..., up to the end of day 5. After ...
- 2,492
0
votes
0
answers
31
views
Churn Risk Modeling without ML
Is it readily possible to do predictive churn analysis (i.e., associating a churn risk with every individual/customer) using statistical tools (e.g. in Excel) not involving the use of machine learning ...
0
votes
1
answer
197
views
How to model a covariate for age in my cox prediction model
I'm currently doing a prediction model using Cox regression on a dataset coming from an ongoing clinical database and containing information about patients who all have the same genetic disease.
In it,...
- 57
0
votes
0
answers
11
views
Predicting yearly cumulative percentage time series
I have a data set about 4G yearly coverage, as a percentage of the total population, it looks like this:
...
- 81
0
votes
0
answers
22
views
Random effects with B-splines: predictions for test data?
I am working with longitudinal measurements of patients, and I want to model the biomarker on the train dataset and then generate predictions for new patients, given that some measurements are already ...
- 1
1
vote
1
answer
60
views
When predicting case status, what is the appropriate way to examine differences in proteomics by gender (/the interaction between protein and gender)? [closed]
I have a dataset I'm working with containing measurements of several thousand proteins per person, with about 150 people split pretty evenly between two groups (cases and controls). Case status is ...
- 25
1
vote
0
answers
15
views
Hypothesis testing a binary predictive model
I would like to perform a hypothesis test to determine if a particular predictive model is a true improvement, or just attributable to randomness. Below is my suggested formulation; the motivation ...
- 121
0
votes
0
answers
12
views
Predicting deterioration of equipment on a production line
Background
There is a production line where there are 10 machines that interact with some tools. The process goes as such.
Machine makes a product
Product is moved into the tool
When the product is ...
- 101
0
votes
1
answer
31
views
How likely is it that our model better than random in the upper corner of the AUC?
We're using forest-based models in a personnel selection context. For a dataset with 57 features, 230 observations, and a binary outcome, we got the following ROC curves.
This shows the first 6 folds ...
- 113
1
vote
1
answer
51
views
Metric or Method to examine the correlation of a ratio
Let $X$, $Y$, $W$, and $Z$ be random variables where $X$ and $Y$ are different metrics of the same data point and $W$ and $Z$ are also different metrics of the same baseline datapoint. The assumption ...
- 121
0
votes
0
answers
137
views
Can we use Variational Mode Decomposition (VMD) on Time-series analysis / prediction?
I am new to time series analysis and signal processing. I would like to ask if it is correct to use a signal decomposition method like Variational Mode Decomposition (VMD) in time series analysis? i.e....
- 1
2
votes
1
answer
146
views
Time series analysis if it is predictable or not
I have two types of time series, stationary time series dataset and non stationary time series dataset. Now I need to know if these two series predictable or not. I read about the models of ...
- 169
1
vote
1
answer
32
views
Choosing best modelling practices
I am modelling some behavioural data in R to determine the best parameters for explaining foraging success of marine animals. I have both fine scale parameters relating to the direct individual (e.g. ...
- 13
1
vote
2
answers
227
views
Prediction/Forecastig of one variable with relation of multiple features
My dataset is composed of time series (40 points) with multiple variables
...
- 13
13
votes
4
answers
4k
views
Is it possible that X and Y are uncorrelated but X can significantly predict Y? [duplicate]
I was just wondering is it possible that X and Y are uncorrelated but X can significantly predict Y?
If so, how would you explain and interpret that?
- 131
2
votes
1
answer
105
views
Model Selection vs. Ensemble Learning
Is model selection just a specific kind of ensemble learning, where ensemble learning is loosely defined as "combining multiple models in some capacity to hopefully get an improved model"?
...
1
vote
1
answer
283
views
Zero-truncated negative binomial model in glmmTMB predictions
I have a dataset of counts of a vocalisation per hour. I am interested in fitting a model to see if the count of the vocalisations of a given category per hour is effected by noise. My response ...
- 11
1
vote
1
answer
68
views
How to find the "best predicting" variable, if the dependent and independent variables are not really correlated?
Here is a short explanation of my problem:
My data:
9 independent variables
1 dependent variable
In theory, the variables should be linear (the higher the independent variable, the higher the ...
- 13
0
votes
0
answers
38
views
Prediction -- conditioning on X vs conditioning on Y?
In general (and perhaps in an overly simplified fashion), we can formalize the goal of using some variable $X$ to predict another variable $Y$ as trying to find a function $f : X \to Y$ which ...
0
votes
0
answers
13
views
What is the name for this type of data and how do I model it?
I have two questions:
What is the name of this type of data
What model is used for prediction?
The data:
You have 100 objects, 30 spheres, 30 cubes, 40 rectangles.
Each object has a weight. Either ...
- 209
2
votes
1
answer
321
views
Prediction metrics for left-truncated and right-censored data
I am trying to assess (out-of-sample) predictive performance of survival analysis models with left-truncated and right-censored data. Assume the training and test data, respectively, consist of ...
- 21
1
vote
1
answer
64
views
Correlated categorical predictors in logistic regression
Suppose I have three categorical variables, x, y, and z and I have a binary response variable r. I want to predict the response using the glm model.
Should I check the association between x, y, and z ...
- 113
2
votes
1
answer
21
views
How can I calculate the optimal formula for predicting an outcome based on a lab value?
How can I calculate the optimal formula for predicting an outcome based on a lab value?
e.g. a rise in liver function tests can be predictive of liver disease.
Having liver disease (or not) is the ...
0
votes
1
answer
71
views
How to avoid estimating prices that are more than 25 dollars off of the actual price in Machine Learning model?
I am currently working on a case study where I have to estimate how much a person makes by giving their property for rent. They provided me with a constraint which is as follows:
"avoid ...
- 109
0
votes
0
answers
26
views
Decision Curve Analysis
I have a question about decision curve analysis. I have trouble understanding the common strategy of treat/intervention to all. I do not understand why the line does not extend towards the whole risk ...
- 15