A data science project that analyzes statistical associations in recipes to create "mean recipes" - statistically-derived recipes based on analysis of thousands of real recipes.
Currently analyzed
9,922 recipes
The Mean Kitchen (TMK) combines quantitative ingredient analysis with pattern-based instruction analysis to generate insights about cooking patterns and recipe optimization.
The project uses a two-pronged approach: analyzing ingredient frequencies and quantities to find statistical patterns, while also examining cooking instructions to identify common techniques and sequences.
TMK has successfully processed and analyzed over 9,922 recipes from the Food.com dataset, extracting statistical insights from 1,350+ unique ingredients and identifying 2,365+ significant ingredient correlations.
Comprehensive analysis tools for recipe data
Statistical analysis of ingredient frequencies, quantities, and correlations across thousands of recipes
Pattern-based instruction analysis to identify common cooking techniques and sequences
Generate mean, median, and mode recipes based on analysis of large recipe datasets
Interactive charts and graphs showing ingredient relationships and cooking patterns
Successfully analyzed 9,922+ recipes from Food.com with 89.2% success rate
Machine learning for outlier detection and recipe recommendations