Tom Tran
- Oct 28, 2021
- 1 min read

World Diets Compared by Machine Learning

World Diet and Cuisine Map by Machine Learning Algorithm

Using clustering algorithm, I created a map based on per capita consumption data including vegetables, seafood, beef, pork, poultry, etc.

The result seems very sensible. The groups are:

1. Mongolia, Gabon and Botswana (possibly other countries without data): The diets in these countries include a lot of goat/mutton in Mongolia, bush meats like antelope, wild boar and monkey in Gabon, and mopane worms in Botswana.

2. USA, Australia, Brazil, etc.: The obvious burger, steak and fries with a lot of beef, diary and potato. Overall, very high meat consumption of most kinds.

3. India, Pakistan, most African countries like Nigeria, Ethiopia, etc.: Low meat, dairy and seafood consumption of all kinds and higher in consumption of vegetables.

4. East and Southeast Asia: A lot of rice and seafood consumption.

5. Canada and most of Europe: Overall, very high meat consumption of most kinds, especially pork. Also high on dairy products and potato.

6. Central and South America, the Caribbean countries, a few Middle Eastern and European nations, Northern African countries: High poultry (e.g. chicken) consumption, quite balanced otherwise.

Below are a few maps of the food consumptions per capita.

Argentina has the world's highest beef consumption per capita.