JFJacob Gates Foster

Jacob Gates Foster

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Jacob Gates Foster has a total of 29 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • A Primer on Deep Learning for Causal Inference

    by Bernard J. Koch, Tim Sainburg, Pablo Geraldo Bastías, Song Jiang, Yizhou Sun, Jacob G. Foster

    This primer systematizes the emerging literature on causal inference using deep neural networks under the potential outcomes framework. It provides an intuitive introduction to building and optimizing custom deep learning models and shows how to adapt them to estimate/predict heterogeneous treatment effects. It also discusses ongoing work to extend causal inference to settings where confounding is nonlinear, time-varying, or encoded in text, networks, and images. To maximize accessibility, we also introduce prerequisite concepts from causal inference and deep learning. The primer differs from other treatments of deep learning and causal inference in its sharp focus on observational causal estimation, its extended exposition of key algorithms, and its detailed tutorials for implementing, training, and selecting among deep estimators in TensorFlow 2 and PyTorch.


  • Open Access

    Identifying Witnessed Suicides in National Violent Death Reporting System Narratives

    by Vickie M. Mays, Mikaela Gareeb, Xingruo Zhang, Vivian Nguyen, Joelle Rosenberg, Yuri Lin, Alina Arseniev-Koehler, Adam Eliav, Jacob Gates Foster, Mika Baumgardner, et al.

    There is increasing attention to suicides that occur in view of others, as these deaths can cause significant psychological impact on witnesses. This study illuminates characteristics of witnessed suicides and compares characteristics of these deaths to non-witnessed suicides. We develop a codable definition of what constitutes witnessed (vs. non-witnessed) suicide. Our data include a sample of 1200 suicide descriptions from the 2003–2017 National Violent Death Reporting System (NVDRS). We first developed criteria to identify probable cases of witnessed suicide. The coding scheme achieved 94.5% agreement and identified approximately 10% (n = 125) of suicides as witnessed. Next, we examined differences between witnessed and non-witnessed suicides in demographics, manner of death, and social/environmental factors using bivariate Chi-squared tests, multivariate logistic regression, and ANOVA. Witnessed suicide decedents were significantly more likely than non-witnessed suicide decedents to be male, younger, and members of a sexual minority, and to have died in living spaces by means of a firearm. Two thirds of witnesses were strangers to the decedents, while 23.2% were romantic partners or ex-partners of the decedents. Our coding method offers a reliable approach to identify witnessed suicides. While witnessed suicides are relatively infrequent, these deaths have profound impact on witnesses. Articulating the features of witnessed suicides may contribute to identifying potential risk mitigation strategies.


  • Open Access

    Machine Learning as a Model for Cultural Learning: Teaching an Algorithm What it Means to be Fat

    by Alina Arseniev-Koehler, Jacob G. Foster

    Public culture is a powerful source of cognitive socialization; for example, media language is full of meanings about body weight. Yet it remains unclear how individuals process meanings in public culture. We suggest that schema learning is a core mechanism by which public culture becomes personal culture. We propose that a burgeoning approach in computational text analysis – neural word embeddings – can be interpreted as a formal model for cultural learning. Embeddings allow us to empirically model schema learning and activation from natural language data. We illustrate our approach by extracting four lower-order schemas from news articles: the gender, moral, health, and class meanings of body weight. Using these lower-order schemas we quantify how words about body weight “fill in the blanks” about gender, morality, health, and class. Our findings reinforce ongoing concerns that machine-learning models (e.g., of natural language) can encode and reproduce harmful human biases.


  • Open Access

    Integrating topic modeling and word embedding to characterize violent deaths

    by Alina Arseniev-Koehler, Susan D. Cochran, Vickie M. Mays, Kai-Wei Chang, Jacob G. Foster

    Significance: We introduce an approach to identify latent topics in large-scale text data. Our approach integrates two prominent methods of computational text analysis: topic modeling and word embedding. We apply our approach to written narratives of violent death (e.g., suicides and homicides) in the National Violent Death Reporting System (NVDRS). Many of our topics reveal aspects of violent death not captured in existing classification schemes. We also extract gender bias in the topics themselves (e.g., a topic about long guns is particularly masculine). Our findings suggest new lines of research that could contribute to reducing suicides or homicides. Our methods are broadly applicable to text data and can unlock similar information in other administrative databases.


  • Top Cited Papers
    Open Access

    Not by transmission alone: the role of invention in cultural evolution

    by Susan Perry, Alecia Carter, Marco Smolla, Erol Akçay, Sabine Nöbel, Jacob G. Foster, Susan D. Healy

    Innovation—the combination of invention and social learning—can empower species to invade new niches via cultural adaptation. Social learning has typically been regarded as the fundamental driver for the emergence of traditions and thus culture. Consequently, invention has been relatively understudied outside the human lineage—despite being the source of new traditions. This neglect leaves basic questions unanswered: what factors promote the creation of new ideas and practices? What affects their spread or loss? We critically review the existing literature, focusing on four levels of investigation: traits (what sorts of behaviours are easiest to invent?), individuals (what factors make some individuals more likely to be inventors?), ecological contexts (what aspects of the environment make invention or transmission more likely?), and populations (what features of relationships and societies promote the rise and spread of new inventions?). We aim to inspire new research by highlighting theoretical and empirical gaps in the study of innovation, focusing primarily on inventions in non-humans. Understanding the role of invention and innovation in the history of life requires a well-developed theoretical framework (which embraces cognitive processes) and a taxonomically broad, cross-species dataset that explicitly investigates inventions and their transmission. We outline such an agenda here. This article is part of the theme issue ‘Foundations of cultural evolution’.


  • Open Access

    Machine learning as a model for cultural learning: Teaching an algorithm what it means to be fat

    by Alina Arseniev-Koehler, Jacob G. Foster

    As we navigate our cultural environment, we learn cultural biases, like those around gender, social class, health, and body weight. It is unclear, however, exactly how public culture becomes private culture. In this paper, we provide a theoretical account of such cultural learning. We propose that neural word embeddings provide a parsimonious and cognitively plausible model of the representations learned from natural language. Using neural word embeddings, we extract cultural schemata about body weight from New York Times articles. We identify several cultural schemata that link obesity to gender, immorality, poor health, and low socioeconomic class. Such schemata may be subtly but pervasively activated in public culture; thus, language can chronically reproduce biases. Our findings reinforce ongoing concerns that machine learning can also encode, and reproduce, harmful human biases.

  • Top Cited Papers

    Tradition and Innovation in Scientists’ Research Strategies

    by Jacob G. Foster, Andrey Rzhetsky, James A. Evans

    What factors affect a scientist’s choice of research problem? Qualitative research in the history and sociology of science suggests that this choice is patterned by an “essential tension” between productive tradition and risky innovation. We examine this tension through Bourdieu’s field theory of science, and we explore it empirically by analyzing millions of biomedical abstracts from MEDLINE. We represent the evolving state of chemical knowledge with networks extracted from these abstracts. We then develop a typology of research strategies on these networks. Scientists can introduce novel chemicals and chemical relationships (innovation) or delve deeper into known ones (tradition). They can consolidate knowledge clusters or bridge them. The aggregate distribution of published strategies remains remarkably stable. High-risk innovation strategies are rare and reflect a growing focus on established knowledge. An innovative publication is more likely to achieve high impact than a conservative one, but the additional reward does not compensate for the risk of failing to publish. By studying prizewinners in biomedicine and chemistry, we show that occasional gambles for extraordinary impact are a compelling explanation for observed levels of risky innovation. Our analysis of the essential tension identifies institutional forces that sustain tradition and suggests policy interventions to foster innovation.


  • Dynamic landscapes: A model of context and contingency in evolution

    by David V. Foster, Mary M. Rorick, Tanja Gesell, Laura M. Feeney, Jacob G. Foster