The value of Bayesian predictive projection for variable selection: An example selecting lifestyle predictors of young adult well-being

Abstract

BackgroundVariable selection is an important issue in many fields such as public health and psychology. Researchers often gather data on many variables of interest and then are faced with two challenging goals: building an accurate model with few predictors, and making probabilistic statements (inference) about this model. Unfortunately, it is currently difficult to attain these goals with the two most popular methods for variable selection methods: stepwise selection and LASSO. The aim of the present study was to demonstrate the use predictive projection feature selection – a novel Bayesian variable selection method that delivers both predictive power and inference. We apply predictive projection to a sample of New Zealand young adults, use it to build a compact model for predicting well-being, and compare it to other variable selection methods. MethodsThe sample consisted of 791 young adults (ages 18 to 25, 71.7% female) from New Zealand who had taken part in the Dunedin Daily Life Study in 2013-2014. Participants completed a 13-day online daily diary assessment of their well-being and a range of lifestyle variables (e.g., sleep, physical activity, diet variables). The participants’ diary data was averaged across days and analyzed cross-sectionally to identify predictors of average flourishing. Predictive projection was used to select as few predictors as necessary to approximate the predictive accuracy of a reference model with all 28 predictors. Predictive projection was also compared to other variable selection methods, including stepwise selection and LASSO.ResultsThree predictors were sufficient to approximate the predictions of the reference model: higher sleep quality, less trouble concentrating, and more servings of fruit. The performance of the projected submodel generalized well. Compared to other variable selection methods, predictive projection lead to models with either matching or slightly worse performance but with much fewer predictors.ConclusionPredictive projection was used to efficiently arrive at a compact model with good predictive accuracy. The predictors selected into the submodel – felt refreshed after waking up, had less trouble concentrating, and ate more servings of fruit – were all theoretically meaningful. Our findings have important implications for applications of variable selection in health research.

Predictors of positive mental health in higher education students. A cross‐sectional predictive study

“…Some variables seem to be students' psychological well‐being predictors, like male gender, dating relationship, good academic performance, exercise, sleeping 7 h per night, and academic life satisfaction (Nogueira & Sequeira, 2020). Evidence shows that exercise, the quality of sleep, and a healthy diet are strong predictors of HES well‐being (Bartonicek et al, 2020; Collins et al, 2020; Mendes et al, 2019). Good night's sleep is essential for a person's well‐functioning.…”

Mixed Effects Machine Learning Models for Colon Cancer Metastasis Prediction using Spatially Localized Immuno-Oncology Markers

“…METS predictions were made within-batch and are not an accurate reflection of deployment on new samples. Projective predictive feature selection can help select more relevant BGLMM predictors 42,43 . We also noticed that there was no gain in performance from using the ROI spatial coordinates, which may have been an artifact of implementational difficulties, the low sample size of ROIs and/or selection/learning of improper kernels.…”

“…We did not incorporate time from initial biopsy to development of local or distant METS, though we plan to comment on the temporality of such associations in follow-up clinical findings. Projective predictive feature selection can help select more relevant BGLMM predictors 42,43 . We neither performed nor compared findings from univariable analyses with multivariable analyses, which is outside the study scope.…”

Identification of Spatial Proteomic Signatures of Colon Tumor Metastasis: A Digital Spatial Profiling Approach

“…Interactions were applied to a Bayesian hierarchical logistic regression model to report pertinent effect modifiers (e.g., effect of CD20, conditional on age; interactions encapsulated in 𝑥 ⃗) 27,37 : 𝑙𝑜𝑔𝑖𝑡(𝑝 " ) = 𝛽 ⃗ ⋅ 𝑥 ⃗ + 𝜃 <+,=> ["] . As many interactions were initially selected, features were selected using the Horseshoe LASSO method and the remaining features were fit with weakly informative priors 38,39 . Effect estimates for salient effect modifiers were reported similar to the previous sections.…”