This project develops a machine learning model to classify the quality of unilateral dumbbell biceps curls based on accelerometer data. Using a dataset of 19,622 observations, we compared classification approaches and found that a Random Forest model provided the highest predictive accuracy (>99%). This report details the data preprocessing steps, feature selection rationale, and model validation results.

Pitch as part of the Coursera Developing Data Products course.

A simple page explaining linear regression made for the "developing data products" Coursera course.