Dataset and Experimental Results for Automatic Timesheet Decomposition: A Comparative Study of Machine Learning Models and Rule-Based Engine

This dataset includes input data, training/test splits, and detailed experimental results from the study “Automatic Timesheet Decomposition: Comparative Approaches Using Machine Learning and Rule-Based Engine”.The research compares predictive models (Decision Tree, Random Forest, Deep Neural Networks) and a self-learning rule engine, evaluating accuracy and generalization on 100 anonymized real-world datasets.Contents: Normalized and quantized datasets (30-minute granularity) Results and accuracy metrics (FullMatch, single-causal)