### README

## Installation

# required packages (tested with R 4.3.3):

caret
kernlab
kknn
plyr
pracma
rpart
rpart.plot
xgboost
cluster
aricode

# R environment

> install.packages("myPackage", dependencies=TRUE)

# alternative (Ubuntu Linux) to install "caret"

> sudo apt-get update
> sudo apt-get install r-cran-caret

## Scripts

emotion_classification.R : emotion classification
intonation_clustering.R : clustering of intonation contour classes
vowel_clustering.R : clustering of vowels

## Data

in data/ subfolder

emodb_feat.csv : Emotion features (eGeMAPS)
emodb_labels.csv : Emotion labels
intonation.csv : polynomial coefficients from f0 contour stylization
vowels.csv : acoustic vowel features and labels

### Links

- Tutorial, list of available models

https://topepo.github.io/caret/

- emodb emotion dataset download and feature extraction

http://blog.syntheticspeech.de/2021/03/07/machine-classification-of-emotional-speech-with-emodb-and-python/

- eGeMAPS acoustic feature set

https://sail.usc.edu/publications/files/eyben-preprinttaffc-2015.pdf