2026 July-August

" Can Yelp become Restaurant Health Inspection Predictor? A Time-Ordered Study of 7,480 Yelp Reviews and 465 Matched Listings" Using grouped cross-validation to compare star ratings, review history, sanitation keywords, and TF-IDF text. The study finds that review language does not reliably predict New York City restaurant health scores better than simple ratings and review counts.

​

2026 April-June

"When More Sugar Makes Less Ethanol: Testing a Four-State Fermentation Model on Public Longan-Waste Data " Here is a yeast paradox: why can more sugar sometimes make less ethanol? Its model shows that highly concentrated mixtures can overwhelm the yeast, leaving it surrounded by sugar but too stressed to turn much of it into ethanol.

​

2026 March-April

"Even Pace Is Not Even Effort: A Grade-Aware Model to Help Run Faster on Hilly Courses" Should runners stubbornly hold one pace on a hilly 10K? The model says no: ease up on climbs and use the downhills, potentially saving a huge amount of seconds without increasing total effort.

​

2025 August-December

"Perfecting Steak Searing with Computation: Heat, Moisture, Flipping, and the Maillard Reaction" This research asks how to build a beautiful steak crust without turning the middle into gray leather. Its model finds that hotter pans and drier surfaces encourage browning, while frequent flipping makes both sides more even—but does not necessarily create the most crust.

​

2025 March-June

"Predicting Ice cream Deliciousness: Composition and Online Ratings of Vanilla Ice Cream" Can a nutrition label can predict deliciousness? After studying 5,793 reviews, it finds that fat, sugar, and long ingredient lists are poor fortune-tellers to icecream enjoyment; creaminess and a pleasant spoonful matter much more.

​

​