Artificial intelligence knows how to calculate life expectancy


AGI – A machine learning approach has been shown to be able to accurately predict several aspects of human life, including the likelihood of premature mortality and personality nuances. The model described in a study published in Nature Computational Scienceyou may be able to provide one quantitative understanding human behavior.

Question regarding the predictability of human life has long been the subject of debate among social scientists. While the sociodemographic factors that play an important role in people’s lives are well known, it has not yet been possible to accurately predict end-of-life consequences.

Using data on the education, health, income, employment and other life events of approximately six million people from the Danish national register, Sune Lehmann, Associate Professor of Social Networks at ETH Zurich, and his colleagues designed a machine learning approach capable of building individual life trajectories. The researchers adapted human language processing techniques to the model.

This allowed them to create a vocabulary that described life events in a similar way to how language models capture complex relationships between words. The proposed model, the so-called ‘life2thing’it creates complex relationships between concepts such as health diagnoses, place of residence, level of income, encoding individual lives using a compact vector representation that forms the basis for predicting life outcomes.

The researchers demonstrated that the model could predict early mortality. Specifically, the model predicted that among the individuals who participated in the study, those sabetween the ages of 35 and 65 would survive 4 years after January 1, 2016. In addition, the model was able to capture personality nuances better than more advanced models, with at least 11% more accuracy.

The results show that by representing the complex connection between social and health data, it is possible to accurately predict life outcomes. However, the authors emphasize that their research is an exploration of what might be possible, and that it should only be used in real-world settings in accordance with rules that protect the rights of individuals.

Reproduction is expressly reserved © Agi 2023


Source link

Leave a Comment