Using sequences of life-events to predict human lives

Research output: Contribution to journal › Journal article › Research › peer-review

Standard

Using sequences of life-events to predict human lives. / Savcisens, Germans; Eliassi-Rad, Tina; Hansen, Lars Kai; Mortensen, Laust Hvas; Lilleholt, Lau; Rogers, Anna; Zettler, Ingo; Lehmann, Sune.

In: Nature Computational Science, Vol. 4, 2024, p. 43–56.

Research output: Contribution to journal › Journal article › Research › peer-review

Harvard

Savcisens, G, Eliassi-Rad, T, Hansen, LK, Mortensen, LH, Lilleholt, L, Rogers, A, Zettler, I & Lehmann, S 2024, 'Using sequences of life-events to predict human lives', Nature Computational Science, vol. 4, pp. 43–56. https://doi.org/10.1038/s43588-023-00573-5

APA

Savcisens, G., Eliassi-Rad, T., Hansen, L. K., Mortensen, L. H., Lilleholt, L., Rogers, A., Zettler, I., & Lehmann, S. (2024). Using sequences of life-events to predict human lives. Nature Computational Science, 4, 43–56. https://doi.org/10.1038/s43588-023-00573-5

Vancouver

Savcisens G, Eliassi-Rad T, Hansen LK, Mortensen LH, Lilleholt L, Rogers A et al. Using sequences of life-events to predict human lives. Nature Computational Science. 2024;4:43–56. https://doi.org/10.1038/s43588-023-00573-5

Author

Savcisens, Germans ; Eliassi-Rad, Tina ; Hansen, Lars Kai ; Mortensen, Laust Hvas ; Lilleholt, Lau ; Rogers, Anna ; Zettler, Ingo ; Lehmann, Sune. / Using sequences of life-events to predict human lives. In: Nature Computational Science. 2024 ; Vol. 4. pp. 43–56.

Bibtex

@article{f6db0cbb8ae946e1bd80e320ace10088,

title = "Using sequences of life-events to predict human lives",

abstract = "Here we represent human lives in a way that shares structural similarity to language, and we exploit this similarity to adapt natural language processing techniques to examine the evolution and predictability of human lives based on detailed event sequences. We do this by drawing on a comprehensive registry dataset, which is available for Denmark across several years, and that includes information about life-events related to health, education, occupation, income, address and working hours, recorded with day-to-day resolution. We create embeddings of life-events in a single vector space, showing that this embedding space is robust and highly structured. Our models allow us to predict diverse outcomes ranging from early mortality to personality nuances, outperforming state-of-the-art models by a wide margin. Using methods for interpreting deep learning models, we probe the algorithm to understand the factors that enable our predictions. Our framework allows researchers to discover potential mechanisms that impact life outcomes as well as the associated possibilities for personalized interventions.",

author = "Germans Savcisens and Tina Eliassi-Rad and Hansen, {Lars Kai} and Mortensen, {Laust Hvas} and Lau Lilleholt and Anna Rogers and Ingo Zettler and Sune Lehmann",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive licence to Springer Nature America, Inc.",

year = "2024",

doi = "10.1038/s43588-023-00573-5",

language = "English",

volume = "4",

pages = "43–56",

journal = "Nature Computational Science",

issn = "2662-8457",

publisher = "Springer",

}

RIS

TY - JOUR

T1 - Using sequences of life-events to predict human lives

AU - Savcisens, Germans

AU - Eliassi-Rad, Tina

AU - Hansen, Lars Kai

AU - Mortensen, Laust Hvas

AU - Lilleholt, Lau

AU - Rogers, Anna

AU - Zettler, Ingo

AU - Lehmann, Sune

PY - 2024

Y1 - 2024

N2 - Here we represent human lives in a way that shares structural similarity to language, and we exploit this similarity to adapt natural language processing techniques to examine the evolution and predictability of human lives based on detailed event sequences. We do this by drawing on a comprehensive registry dataset, which is available for Denmark across several years, and that includes information about life-events related to health, education, occupation, income, address and working hours, recorded with day-to-day resolution. We create embeddings of life-events in a single vector space, showing that this embedding space is robust and highly structured. Our models allow us to predict diverse outcomes ranging from early mortality to personality nuances, outperforming state-of-the-art models by a wide margin. Using methods for interpreting deep learning models, we probe the algorithm to understand the factors that enable our predictions. Our framework allows researchers to discover potential mechanisms that impact life outcomes as well as the associated possibilities for personalized interventions.

AB - Here we represent human lives in a way that shares structural similarity to language, and we exploit this similarity to adapt natural language processing techniques to examine the evolution and predictability of human lives based on detailed event sequences. We do this by drawing on a comprehensive registry dataset, which is available for Denmark across several years, and that includes information about life-events related to health, education, occupation, income, address and working hours, recorded with day-to-day resolution. We create embeddings of life-events in a single vector space, showing that this embedding space is robust and highly structured. Our models allow us to predict diverse outcomes ranging from early mortality to personality nuances, outperforming state-of-the-art models by a wide margin. Using methods for interpreting deep learning models, we probe the algorithm to understand the factors that enable our predictions. Our framework allows researchers to discover potential mechanisms that impact life outcomes as well as the associated possibilities for personalized interventions.

U2 - 10.1038/s43588-023-00573-5

DO - 10.1038/s43588-023-00573-5

M3 - Journal article

C2 - 38177491

AN - SCOPUS:85180261076

VL - 4

SP - 43

EP - 56

JO - Nature Computational Science

JF - Nature Computational Science

SN - 2662-8457

ER -

ID: 378520874

Department of Psychology