| 1. |
- Annadani, Yashas, et al.
(författare)
-
BayesDAG : Gradient-Based Posterior Inference for Causal Discovery
- 2023
-
Ingår i: Advances in Neural Information Processing Systems 36 - 37th Conference on Neural Information Processing Systems, NeurIPS 2023. - : Neural Information Processing Systems Foundation.
-
Konferensbidrag (refereegranskat)abstract
- Bayesian causal discovery aims to infer the posterior distribution over causal models from observed data, quantifying epistemic uncertainty and benefiting downstream tasks. However, computational challenges arise due to joint inference over combinatorial space of Directed Acyclic Graphs (DAGs) and nonlinear functions. Despite recent progress towards efficient posterior inference over DAGs, existing methods are either limited to variational inference on node permutation matrices for linear causal models, leading to compromised inference accuracy, or continuous relaxation of adjacency matrices constrained by a DAG regularizer, which cannot ensure resulting graphs are DAGs. In this work, we introduce a scalable Bayesian causal discovery framework based on a combination of stochastic gradient Markov Chain Monte Carlo (SG-MCMC) and Variational Inference (VI) that overcomes these limitations. Our approach directly samples DAGs from the posterior without requiring any DAG regularization, simultaneously draws function parameter samples and is applicable to both linear and nonlinear causal models. To enable our approach, we derive a novel equivalence to the permutation-based DAG learning, which opens up possibilities of using any relaxed gradient estimator defined over permutations. To our knowledge, this is the first framework applying gradient-based MCMC sampling for causal discovery. Empirical evaluation on synthetic and real-world datasets demonstrate our approach's effectiveness compared to state-of-the-art baselines.
|
|
| 2. |
- Crona, Beatrice, et al.
(författare)
-
China at a Crossroads : An Analysis of China's Changing Seafood Production and Consumption
- 2020
-
Ingår i: One Earth. - : Elsevier BV. - 2590-3330 .- 2590-3322. ; 3:1, s. 32-44
-
Tidskriftsartikel (refereegranskat)abstract
- China is a key player in global production, consumption, and trade of seafood. Given this dominance, Chinese choices regarding what seafood to eat, and how and where to source it, are increasingly important—for China, and for the rest of the world. This perspective explores this issue using a transdisciplinary approach and discusses plausible trajectories and implications for assumptions of future modeling efforts and global environmental sustainability and seafood supply. We outline China's 2030 projected domestic seafood production and consumption through an examination of available statistics, and qualitatively evaluate these in relation to key stated Chinese policy targets, consumer trends, and dominant political narratives. Our analysis shows that by 2030 China is likely to see seafood consumption outstrip domestic production. To meet the seafood gap China will likely attempt to increase domestic freshwater and offshore aquaculture, increase seafood imports, possibly expand the distant water fishing industry, and invest in seafood production abroad.
|
|
| 3. |
- Gephart, Jessica, et al.
(författare)
-
Environmental performance of blue foods
- 2021
-
Ingår i: Nature. - : Nature Research. - 0028-0836 .- 1476-4687. ; 597:7876, s. 360-365
-
Tidskriftsartikel (refereegranskat)abstract
- Fish and other aquatic foods (blue foods) present an opportunity for more sustainable diets1,2. Yet comprehensive comparison has been limited due to sparse inclusion of blue foods in environmental impact studies3,4 relative to the vast diversity of production5. Here we provide standardized estimates of greenhouse gas, nitrogen, phosphorus, freshwater and land stressors for species groups covering nearly three quarters of global production. We find that across all blue foods, farmed bivalves and seaweeds generate the lowest stressors. Capture fisheries predominantly generate greenhouse gas emissions, with small pelagic fishes generating lower emissions than all fed aquaculture, but flatfish and crustaceans generating the highest. Among farmed finfish and crustaceans, silver and bighead carps have the lowest greenhouse gas, nitrogen and phosphorus emissions, but highest water use, while farmed salmon and trout use the least land and water. Finally, we model intervention scenarios and find improving feed conversion ratios reduces stressors across all fed groups, increasing fish yield reduces land and water use by up to half, and optimizing gears reduces capture fishery emissions by more than half for some groups. Collectively, our analysis identifies high-performing blue foods, highlights opportunities to improve environmental performance, advances data-poor environmental assessments, and informs sustainable diets. © 2021, The Author(s)
|
|
| 4. |
- Li, Sirui, et al.
(författare)
-
Glioma grading, molecular feature classification, and microstructural characterization using MR diffusional variance decomposition (DIVIDE) imaging
- 2021
-
Ingår i: European Radiology. - : Springer Science and Business Media LLC. - 0938-7994 .- 1432-1084. ; 31:11, s. 8197-8207
-
Tidskriftsartikel (refereegranskat)abstract
- Objective: To evaluate the potential of diffusional variance decomposition (DIVIDE) for grading, molecular feature classification, and microstructural characterization of gliomas. Materials and methods: Participants with suspected gliomas underwent DIVIDE imaging, yielding parameter maps of fractional anisotropy (FA), mean diffusivity (MD), anisotropic mean kurtosis (MKA), isotropic mean kurtosis (MKI), total mean kurtosis (MKT), MKA/MKT, and microscopic fractional anisotropy (μFA). Tumor type and grade, isocitrate dehydrogenase (IDH) 1/2 mutant status, and the Ki-67 labeling index (Ki-67 LI) were determined after surgery. Statistical analysis included 33 high-grade gliomas (HGG) and 17 low-grade gliomas (LGG). Tumor diffusion metrics were compared between HGG and LGG, among grades, and between wild and mutated IDH types using appropriate tests according to normality assessment results. Receiver operating characteristic and Spearman correlation analysis were also used for statistical evaluations. Results: FA, MD, MKA, MKI, MKT, μFA, and MKA/MKT differed between HGG and LGG (FA: p = 0.047; MD: p = 0.037, others p < 0.001), and among glioma grade II, III, and IV (FA: p = 0.048; MD: p = 0.038, others p < 0.001). All diffusion metrics differed between wild-type and mutated IDH tumors (MKI: p = 0.003; others: p < 0.001). The metrics that best discriminated between HGG and LGGs and between wild-type and mutated IDH tumors were MKT and FA respectively (area under the curve 0.866 and 0.881). All diffusion metrics except FA showed significant correlation with Ki-67 LI, and MKI had the highest correlation coefficient (rs = 0.618). Conclusion: DIVIDE is a promising technique for glioma characterization and diagnosis. Key Points: • DIVIDE metrics MKIis related to cell density heterogeneity while MKAand μFA are related to cell eccentricity. • DIVIDE metrics can effectively differentiate LGG from HGG and IDH mutation from wild-type tumor, and showed significant correlation with the Ki-67 labeling index. • MKIwas larger than MKAwhich indicates predominant cell density heterogeneity in gliomas. • MKAand MKIincreased with grade or degree of malignancy, however with a relatively larger increase in the cell eccentricity metric MKAin relation to the cell density heterogeneity metric MKI.
|
|
| 5. |
- Love, David C., et al.
(författare)
-
Emerging COVID-19 impacts, responses, and lessons for building resilience in the seafood system
- 2021
-
Ingår i: Global food security. - : Elsevier BV. - 2211-9124. ; 28
-
Tidskriftsartikel (refereegranskat)abstract
- The COVID-19 pandemic and subsequent lockdowns are creating health and economic crises that threaten food and nutrition security. The seafood sector provides important sources of nutrition and employment, especially in low-income countries, and is highly globalized allowing shocks to propagate. We studied COVID-19-related disruptions, impacts, and responses to the seafood sector from January through May 2020, using a food system resilience ‘action cycle’ framework as a guide. We find that some supply chains, market segments, companies, small-scale actors and civil society have shown initial signs of greater resilience than others. COVID-19 has also highlighted the vulnerability of certain groups working in- or dependent on the seafood sector. We discuss early coping and adaptive responses combined with lessons from past shocks that could be considered when building resilience in the sector. We end with strategic research needs to support learning from COVID-19 impacts and responses.
|
|
| 6. |
- Pan, Wenbo, et al.
(författare)
-
Earthquake-induced fracture displacements and transmissivity changes in a 3D fracture network of crystalline rock for spent nuclear fuel disposal
- 2023
-
Ingår i: Journal of Rock Mechanics and Geotechnical Engineering. - : Elsevier. - 1674-7755. ; 15:9, s. 2313-2329
-
Tidskriftsartikel (refereegranskat)abstract
- During the long service period of a nuclear waste repository in crystalline rock, large earthquake(s) may occur nearby the repository site and coseismically alter the local stress field around pre-existing fractures within the geological formation. The resulting fracture normal/shear displacements may lead to fracture opening and further promote the transport of leaked radionuclides into the groundwater system. Thus, it is of central importance to analyze the consequences of potential future earthquake(s) on the hydrogeological properties of a repository site for spent nuclear fuel disposal. Based on the detailed site characterization data of the repository site at Forsmark, Sweden, we conduct a three-dimensional (3D) seismo-hydro-mechanical simulation using the 3Dimensional Distinct Element Code (3DEC). We explicitly represent a primary seismogenic fault zone and its surrounding secondary fracture network associated with a power-law size scaling and a Fisher orientation distribution. An earthquake with a magnitude of Mw = 5.6 caused by the reactivation of the primary fault zone is modeled by simulating its transient rupture propagating radially outwards from a predefined hypocenter at a specified rupture speed, with the faulting dynamics controlled by a strength weakening law. We model the coseismic response of the off-fault fracture network subject to both static and dynamic triggering effects. We further diagnose the distribution of fracture hydro-mechanical properties (e.g. mechanical/hydraulic aperture, hydraulic transmissivity) before and after the earthquake in order to quantify earthquake-induced hydraulic changes in the fracture network. It is found that earthquake-induced fracture transmissivity changes tend to follow a power-law decay with the distance to the earthquake fault. Our simulation results and insights obtained have important implications for the long-term performance assessment of nuclear waste repositories in fractured crystalline rocks.
|
|
| 7. |
|
|
| 8. |
- Short, Rebecca E., et al.
(författare)
-
Harnessing the diversity of small-scale actors is key to the future of aquatic food systems
- 2021
-
Ingår i: Nature Food. - : Springer Science and Business Media LLC. - 2662-1355. ; 2:9, s. 733-741
-
Tidskriftsartikel (refereegranskat)abstract
- Small-scale fisheries and aquaculture (SSFA) provide livelihoods for over 100 million people and sustenance for -1 billion people, particularly in the Global South. Aquatic foods are distributed through diverse supply chains, with the potential to be highly adaptable to stresses and shocks, but face a growing range of threats and adaptive challenges. Contemporary governance assumes homogeneity in SSFA despite the diverse nature of this sector. Here we use SSFA actor profiles to capture the key dimensions and dynamism of SSFA diversity, reviewing contemporary threats and exploring opportunities for the SSFA sector. The heuristic framework can inform adaptive governance actions supporting the diversity and vital roles of SSFA in food systems, and in the health and livelihoods of nutritionally vulnerable people-supporting their viability through appropriate policies whilst fostering equitable and sustainable food systems.
|
|
| 9. |
- Tan, Liping, et al.
(författare)
-
Porous fibrous bacterial cellulose/La(OH)3 membrane for superior phosphate removal from water
- 2022
-
Ingår i: Carbohydrate Polymers. - : Elsevier BV. - 0144-8617 .- 1879-1344. ; 298
-
Tidskriftsartikel (refereegranskat)abstract
- Lanthanum (La)-based nanoparticles (NPs) are promising candidates for phosphate removal owing to their inherently high affinity towards phosphate. However, significant challenges remain to be addressed before their practical deployment, especially the problems associated with their aggregation. Herein, we fabricated a high-efficient sorbent for phosphate removal through in-situ synthesizing La(OH)3 NPs on a natural support, bacterial cellulose (BC), which is pre-modified with polyethyleneimine. The resultant La(OH)3 NPs-immobilized BC with different La contents (BPLa-X) exhibited a highly fibrous porous structure, in which BPLa-3 was selected for further phosphate adsorption studies. BPLa-3 demonstrated a high adsorption capacity of 125.5 mg P g−1, and high adsorption selectivity due to the large surface area and abundant exposed active adsorption sites for phosphate. Additionally, BPLa-3 also displayed high reusability and still possessed high adsorption capacity after four consecutive cycles of adsorption-desorption. Therefore, the present adsorbent is believed to be a promising candidate for practical phosphate removal.
|
|
| 10. |
- Wu, Wenbo, et al.
(författare)
-
Deep discriminative clustering and structural constraint for cross-domain fault diagnosis of rotating machinery
- 2023
-
Ingår i: Manufacturing Letters. - : Elsevier BV. - 2213-8463. ; 35, s. 1072-1080
-
Tidskriftsartikel (refereegranskat)abstract
- With the rapid development of intelligent manufacturing, fault diagnostic methods based on deep learning have achieved impressive results. However, most methods require plentiful annotated samples and are based on the assumption that data from the source and target domains has the same distribution. These two conditions are difficult to satisfy in practical engineering. In light of these problems, an unsupervised domain adaptation approach named Deep Discriminative Clustering network with Structural Constraint (DDCSC) is proposed in this article. In our method, a Convolutional Neural Network (CNN) module is exploited for learning feature representations of raw data. Then a softmax module is employed to simultaneously predict class probabilities and cluster assignments of the source and target data, respectively. The learnable cluster centroids are introduced into the latent feature space to alleviate the data distribution discrepancy while better capturing the discriminative structure of the target data. In addition, geometric properties of the source data in a feature space are constrained to expand the scope of each category, which facilitates to improve prediction accuracy. An information-theoretic metric is considered as the objective function of discriminative clustering. Diagnostic experiments on a rolling bearing dataset demonstrate that our approach outperforms other popular intelligent approaches and confirms the effectiveness of discriminative clustering.
|
|
