ECK
Back to Projects
data

MATLAB Code vs App: Emotion Analysis

A comprehensive comparison between manually coded MATLAB models and models generated using MATLAB's Classification Learner app for emotion analysis. Tests Decision Tree classifiers on a 52,681-sample dataset with 14,414 TF-IDF features, finding the App is ~4.7x faster with identical accuracy (~65%).

Tech Stack

MATLABTF-IDFDecision TreeMachine LearningNLP

Key Features

0152,681 samples with 14,414 TF-IDF features
025-fold cross-validation comparison
03App is ~4.7x faster with identical ~65% accuracy
04Structured development pipeline (loading, preprocessing, training)