Prepping the Bar: Insights from Cleaning & Labeling a Cocktail Dataset
summary
Cleaned a noisy cocktail dataset, normalized brand and ingredient names, split citrus into juice vs spirit, and fixed swapped measurement/ingredient fields—catching obvious errors like bitters flagged as 1 oz errors. Extracted features (ABV fraction, bubbles, citrus, sweetness, glass size), encoded simple heuristics to label seasonality (summer tall/low-ABV, winter short/high-ABV), then uses clustering to fill missing labels. Ends with a working inventory-planning prototype driven by those labels.
AI-written summary · about
chapters (5)
- 0:00Start
- 5:25Domain Expertise
- 6:54Data Validation
- 7:49Feature Vectors
- 12:42The Importance of the Domain Expertise
related talks
31mFundamentals of Data Visualization
24mThe TL;DR of EDA
18mData Edibilization: Representing Data with Food
30mData-Centric AI Cookbook-let's prep that data
36mIn the Name of Whiskey
47mQuick and Dirty Data Science with Incanter
31mHoly D@t*! How to Deal with Imperfect, Unclean Datasets
34mWhisky Clustering with Apache Projects: Groovy, Commons Math, Ignite, Spark, Wayang, Beam and Flink