Based on a study by author and Macquarie University psychology researcher Dr Ian Stephen, a computer model has now been trained to recognise a prospective “attractive” person.

Attraction between two people may not just be the result of toned abs, the shape of the face, or curves. Dr Stephen’s finding add to the increasingly popular theory which suggests that modern-day attraction has less to do with personal taste and more to do with our biological instincts.

Based on Dr Stephen’s research, the computer model was programmed to recognise the “healthiest” face on the basis of the three major indicators of good health: body mass index (BMI), body fat percentage and blood pressure.

The computer was reportedly trained to recognise these three health indicators for 272 Asian, African and Caucasian faces. Once trained, the model was then able to accurately predict the BMI, body fat and blood pressure of any individual, just by analysing an image of their face.

“There’s always been this ‘beauty is in the eye of the beholder’ idea which essentially says that what we find attractive is just arbitrary and there is nothing inherent about an attractive person… Evolutionary theory has chipped in on that and essentially says that attractiveness is essentially a mechanism for recognising healthy, fertile, appropriate mates and also healthy friends and allies, because there’s obvious advantages to being friends with or mating with people who are healthy,” Dr Stephen said.

Dr Stephen was also part of a team who found that men who eat a diet rich in fruit and vegetables smell more attractive to women than those who eat high-carbohydrate or high-fat diets.