An official website of the United States government
Alzheimer’s disease (AD) is the most prevalent neurodegenerative disease, yet its current treatments are limited to stopping disease progression. Moreover, the effectiveness of these treatments remains uncertain due to the heterogeneity of the disease. Therefore, it is essential to identify disease subtypes at a very early stage. Current data-driven approaches can be used to classify subtypes during later stages of AD or related disorders, but making predictions in the asymptomatic or prodromal stage is challenging. Furthermore, the classifications of most existing models lack explainability, and these models rely solely on a single modality for assessment, limiting the scope of their analysis. Thus, we propose a multimodal framework that utilizes early-stage indicators, including imaging, genetics, and clinical assessments, to classify AD patients into progression-specific subtypes at an early stage. In our framework, we introduce a tri-modal co-attention mechanism (Tri-COAT) to explicitly capture cross-modal feature associations. Data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (slow progressing = 177, intermediate = 302, and fast = 15) were used to train and evaluate Tri-COAT using a 10-fold stratified cross-testing approach. Our proposed model outperforms baseline models and sheds light on essential associations across multimodal features supported by known biological mechanisms. The multimodal design behind Tri-COAT allows it to achieve the highest classification area under the receiver operating characteristic curve while simultaneously providing interpretability to the model predictions through the co-attention mechanism.
more »
« less
A practical introduction to the Rational Speech Act modeling framework
Recent advances in computational cognitive science (i.e., simulation- based probabilistic programs) have paved the way for significant progress in formal, implementable models of pragmatics. Rather than describing a pragmatic reasoning process in prose, these models formalize and implement one, deriving both qualitative and quantitative predictions of human behavior—predictions that consistently prove correct, demonstrating the viability and value of the framework. The current paper provides a practical introduction to and critical assessment of the Bayesian Rational Speech Act modeling framework, unpacking theoretical foundations, exploring technological innovations, and drawing connections to issues beyond current applications.
more »
« less
- Award ID(s):
- 1911790
- PAR ID:
- 10279450
- Date Published:
- Journal Name:
- ArXivorg
- ISSN:
- 2331-8422
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Climate change and unsustainable land management practices have resulted in extensive soil degradation, including alteration of soil structure (i.e., aggregate and pore size distributions), loss of soil organic carbon, and reduction of water and nutrient holding capacities. Although soil structure, hydrologic processes, and biogeochemical fluxes are tightly linked, their interaction is often unaccounted for in current ecohydrological, hydrological and terrestrial biosphere models. For more holistic predictions of soil hydrological and biogeochemical cycles, models need to incorporate soil structure and macroporosity dynamics, whether in a natural or agricultural ecosystem. Here, we present a theoretical framework that couples soil hydrologic processes and soil microbial activity to soil organic carbon dynamics through the dynamics of soil structure. In particular, we link the Millennial model for soil carbon dynamics, which explicitly models the formation and breakdown of soil aggregates, to a recent parameterization of the soil water retention and hydraulic conductivity curves and to solute and O2diffusivities to soil microsites based on soil macroporosity. To illustrate the significance of incorporating the dynamics of soil structure, we apply the framework to a case study in which soil and vegetation recover over time from agricultural practices. The new framework enables more holistic predictions of the effects of climate change and land management practices on coupled soil hydrological and biogeochemical cycles.more » « less
-
Comparing magnetopause predictions from two MHD models during a geomagnetic storm and a quiet periodMagnetopause location is an important prediction of numerical simulations of the magnetosphere, yet the models can err, either under-predicting or over-predicting the motion of the boundary. This study compares results from two of the most widely used magnetohydrodynamic (MHD) models, the Lyon–Fedder–Mobarry (LFM) model and the Space Weather Modeling Framework (SWMF), to data from the GOES 13 and 15 satellites during the geomagnetic storm on 22 June 2015, and to THEMIS A, D, and E during a quiet period on 31 January 2013. The models not only reproduce the magnetopause crossings of the spacecraft during the storm, but they also predict spurious magnetopause motion after the crossings seen in the GOES data. We investigate the possible causes of the over-predictions during the storm and find the following. First, using different ionospheric conductance models does not significantly alter predictions of the magnetopause location. Second, coupling the Rice Convection Model (RCM) to the MHD codes improves the SWMF magnetopause predictions more than it does for the LFM predictions. Third, the SWMF produces a stronger ring current than LFM, both with and without the RCM and regardless of the LFM spatial resolution. During the non-storm event, LFM predicts the THEMIS magnetopause crossings due to the southward interplanetary magnetic field better than the SWMF. Additionally, increasing the LFM spatial grid resolution improves the THEMIS predictions, while increasing the SWMF grid resolutions does not.more » « less
-
While neural models show remarkable accuracy on individual predictions, their internal beliefs can be inconsistent across examples.In this paper, we formalize such inconsistency as a generalization of prediction error. We propose a learning framework for constraining models using logic rules to regularize them away from inconsistency. Our framework can leverage both labeled and unlabeled examples and is directly compatible with off-the-shelf learning schemes without model redesign. We instantiate our framework on natural language inference, where experiments show that en-forcing invariants stated in logic can help make the predictions of neural models both accurate and consistentmore » « less
-
Summary Forests are a critical carbon sink and widespread tree mortality resulting from climate‐induced drought stress has the potential to alter forests from a carbon sink to a source, causing a positive feedback on climate change. Process‐based vegetation models aim to represent the current understanding of the underlying mechanisms governing plant physiological and ecological responses to climate. Yet model accuracy varies across scales, and regional‐scale model predictive skill is frequently poor when compared with observations of drought‐driven mortality. I propose a framework that leverages differences in model predictive skill across spatial scales, mismatches between model predictions and observations, and differences in the mechanisms included and absent across models to advance the understanding of the physiological and ecological processes driving observed patterns drought‐driven mortality.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- The DOI is not currently available.
