New answers tagged

0

You should build one model. The model should include either deterministic or random term(s) capturing the uniqueness of each week. Example: $$ {\rm Sales\ Revenue} = \beta_0 + \beta_1 * {\rm Time} + \beta_2 * {\rm Time}^2 + ... $$ The complexity of entertained models depends on the sample size. Now, your interest in ridge regression signals your interest in ...


0

If you want the intercept to be the mean of the levels, then effects coding will give you an unweighted mean as your intercept. It will have a standard errors. Moreover, you can now use the standard errors produced by your beta estimates to calculate confidence intervals to see if they cross the unweighted grand mean. It can be done in R using the contrasts ...


0

I am assuming you are building regression model. There is no robust metrics which works for all kind of data. Generally, in regression model you can use MSE, RMSE, MAPE etc. MSE and RMSE are the metrics which can be useful if you are comparing multiple models on same data. But if you do not have multiple models and trying to rely on only one model then you ...


0

I'm not sure from what range I am allowed to say that the model is accurate. Is there any range that is commonly used? What accuracy is "good" will depend on your specific domain. There are no general benchmarks. Best to compare your algorithm to very simple benchmark methods. More at Is there any standard / criteria of good forecast measured by ...


0

I think this chapter could be quite helpful to you: http://peopleanalytics-regression-book.org/survival.html. It focuses specifically on HR-related analyses, which should translate to what you are trying to do here. One clarification that may be helpful: your input data should already tell you which employees left, and when. The goal of survival analysis in ...


0

In the end, you want to end up with the best model possible, the most predictive one regarding brand new data inputs in the future. And, you develop such a model by including all the data you have.


1

Think very carefully about whether you really want to do completely separate models. All regression models run a risk of omitted-variable bias when you omit predictors associated with outcome. With Cox regression, as for logistic regression, this can happen even if the omitted predictor isn't correlated with the included predictors. If your interest is in ...


1

There are a few definitions of baseline model. Typically, it's also the null model, which contains no parameters (except means and variances). It's (just about) the worst model you could have. But that can get more complicated: https://www.researchgate.net/publication/...


0

Say that your data is photos of cars from New York City, including regular cars and yellow cabs. You are building a cab classifier. You split your data randomly to train and test set and assess accuracy. Say that your algorithm has learned to classify all yellow cars as cabs. If the split was random, you would expect similar proportion of yellow cars in ...


Top 50 recent answers are included