This article develops an analysis pipeline for quantifying and relating mouth shape variation to the emotions perceived from facial expressions. We use open-source data that contains ratings from 802 fairgoers on 27 smile-like expressions. Each rater was given a list of seven emotions (happy, sad, anger, contempt, fear, surprise, and disgust) and asked to select all of the words that best described the facial expression. To develop a generalizable method for quantifying mouth shape variation, we leverage statistical shape analysis techniques to parameterize each mouth’s shape in terms of 30 systematically placed landmarks that outline the upper and lower lips. Furthermore, we demonstrate that a three-dimensional representation of these landmark coordinates produces an interpretable feature set that outperforms the original and full-dimensional feature sets in terms of predictive performance. To connect the mouth shape features to the emotion ratings, we develop a nonparametric multinomial regression model that is capable of shrinkage and selection with high-dimensional predictors. Our results demonstrate that the proposed method can produce easily interpretable model predictions that enhance our understanding of the nature in which subtle variations in mouth shape affect the perception of a facial expression.

In proteomics study, Imaging Mass Spectrometry (IMS) is an emerging and very promisingnew technique for protein analysis from intact biological tissues. Though it has showngreat potential and is very promising for rapid mapping of protein localization and thedetection of sizeable differences in protein expression, challenges remain in dataprocessing due to the difficulty of high dimensionality and the fact that the number ofinput variables in prediction model is significantly larger than the number ofobservations. To obtain a complete overview of IMS data and find trace features based onboth spectral and spatial patterns, one faces a global optimization problem. In thispaper, we propose a weighted elastic net (WEN) model based on IMS data processing needs ofusing both the spectral and spatial information for biomarker selection andclassification. Properties including variable selection accuracy of the WEN model arediscussed. Experimental IMS data analysis results show that such a model not only reducesthe number of side features but also helps new biomarkers discovery.