DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information
  1. Fair Concurrent Training of Multiple Models in Federated Learning

    Federated learning (FL) enables collaborative learning across multiple clients. In most FL work, all clients train a single learning task. However, the recent proliferation of FL applications may increasingly require multiple FL tasks to be trained simultaneously, sharing clients’ computing resources, which we call Multiple-Model Federated Learning (MMFL). Current MMFL algorithms use naïve average-based client-task allocation schemes that often lead to unfair performance when FL tasks have heterogeneous difficulty levels, as the more difficult tasks may need more client participation to train effectively. Furthermore, in the MMFL setting, we face a further challenge that some clients may prefer training specificmore » tasks to others, and may not even be willing to train other tasks, e.g., due to high computational costs, which may exacerbate unfairness in training outcomes across tasks. We address both challenges by firstly designing FedFairMMFL, a difficulty-aware algorithm that dynamically allocates clients to tasks in each training round, based on the tasks’ current performance levels. We provide guarantees on the resulting task fairness and FedFairMMFL’s convergence rate. We then propose novel auction designs that incentivizes clients to train multiple tasks, so as to fairly distribute clients’ training efforts across the tasks, and extend our convergence guarantees to this setting. Here, we finally evaluate our algorithm with multiple sets of learning tasks on real world datasets, showing that our algorithm improves fairness by improving the final model accuracy and convergence speed of the worst performing tasks, while maintaining the average accuracy across tasks.« less
  2. Effect of printing parameters on 3D-printed carbon fiber-reinforced polymer composites under magnetic field control

    Magnetic field controlled (MFC) 3D printing is a promising method for realizing the intelligent control of the structural performance of carbon fiber-reinforced plastic (CFRP) composites by adjusting the fiber orientation during the printing process. However, few studies have focused on the influence of the printing parameters on the magnetic control of the fiber direction. Therefore, in this study, the mechanical properties and internal fiber distributions of MFC 3D-printed samples of CFRP composites with 10 wt% fiber content, which were printed at different speeds and layer thicknesses, were investigated. Subsequently, the effects of the fiber content on the mechanical properties andmore » fiber distribution of the MFC 3D-printed samples were studied. Furthermore, the origin of fiber fracture was evaluated through observation and analysis of the fracture section using scanning electron microscopy. The fiber orientation could be well adjusted under the action of a magnetic field and adjustment of printing parameters, thereby improving the tensile strength, young's modulus significantly with slightly decrease of fracture elongation. The recommended parameter combinations (printing speed and layer thickness) were 40 mm/s and 0.2 mm for 10 wt% fiber content CFRP composites.« less
  3. Photodepositing CdS on the Active Cyano Groups Decorated g‐C 3 N 4 in Z‐Scheme Manner Promotes Visible‐Light‐Driven Hydrogen Evolution

    Abstract g‐C 3 N 4 /CdS heterojunctions are potential photocatalysts for hydrogen production but their traditional type‐II configuration generally leads to weak oxidative and reductive activity. How to construct the novel Z‐scheme g‐C 3 N 4 /CdS counterparts to address this issue remains a great challenge in this field. In this work, a new direct Z‐scheme heterojunction of defective g‐C 3 N 4 /CdS is designed by introducing cyano groups (NC‐) as the active bridge sites. Experimental observations in combination with density functional theory (DFT) calculations reveal that the unique electron‐withdrawing feature of cyano groups in the defective g‐C 3more » N 4 /CdS heterostructure can endow this photocatalyst with numerous advantageous properties including high light absorption ability, strong redox performance, satisfactory charge separation efficiency, and long lifetime of charge carriers. Consequently, the resultant photocatalytic system exhibits more active performance than CdS and g‐C 3 N 4 under visible light and reaches an excellent hydrogen evolution rate of 1809.07 µmol h −1 g −1 , which is 6.09 times higher than pristine g‐C 3 N 4 . Moreover, the defective g‐C 3 N 4 /CdS photocatalyst maintains good stability after 40 h continuous test. This work provides new insights into design and construction of Z‐scheme heterojunctions for regulating the visible‐light‐induced photocatalytic activity for H 2 evolution.« less
  4. Co‐culture engineering for microbial biosynthesis of 3‐amino‐benzoic acid in Escherichia coli

    Abstract 3‐amino‐benzoic acid (3AB) is an important building block molecule for production of a wide range of important compounds such as natural products with various biological activities. In the present study, we established a microbial biosynthetic system for de novo 3AB production from the simple substrate glucose. First, the active 3AB biosynthetic pathway was reconstituted in the bacterium Escherichia coli , which resulted in the production of 1.5 mg/L 3AB. In an effort to improve the production, an E. coli ‐ E. coli co‐culture system was engineered to modularize the biosynthetic pathway between an upstream strain and an downstream strain.more » Specifically, the upstream biosynthetic module was contained in a fixed E. coli strain, whereas a series of E. coli strains were engineered to accommodate the downstream biosynthetic module and screened for optimal production performance. The best co‐culture system was found to improve 3AB production by 15 fold, compared to the mono‐culture approach. Further engineering of the co‐culture system resulted in biosynthesis of 48 mg/L 3AB. Our results demonstrate co‐culture engineering can be a powerful new approach in the broad field of metabolic engineering.« less
  5. Engineering lipid overproduction in the oleaginous yeast Yarrowia lipolytica

  6. Engineering Escherichia coli coculture systems for the production of biochemical products

    Significance Production of industrial compounds by using engineered microorganisms is a robust method to reduce our reliance on nonrenewable petroleum resources and increase the utility of renewable resources, such as lignocellulose. Because there are major limitations for engineering a single microbial cell to achieve high-yield production, we developed microbial coculture systems consisting of two different microbial cell types to use sugar mixtures that can be derived from lignocellulose efficiently. We demonstrate that this approach is successful for achieving high-level production of two important value-added molecules, cis , cis -muconic acid and 4-hydroxybenzoic acid. This accomplishment establishes a previously unidentified technologymore » for advancing future research in metabolic engineering and synthetic biology.« less
  7. Engineering a novel biosynthetic pathway in Escherichia coli for production of renewable ethylene glycol

    ABSTRACT Ethylene glycol (EG) is an important commodity chemical with broad industrial applications. It is presently produced from petroleum or natural gas feedstocks in processes requiring consumption of significant quantities of non‐renewable resources. Here, we report a novel pathway for biosynthesis of EG from the renewable sugar glucose in metabolically engineered Escherichia coli . Serine‐to‐EG conversion was first achieved through a pathway comprising serine decarboxylase, ethanolamine oxidase, and glycolaldehyde reductase. Serine provision in E. coli was then enhanced by overexpression of the serine‐biosynthesis pathway. The integration of these two parts into the complete EG‐biosynthesis pathway in E. coli allowed formore » production of 4.1 g/L EG at a cumulative yield of 0.14 g‐EG/g‐glucose, establishing a foundation for a promising biotechnology. Biotechnol. Bioeng. 2016;113: 376–383. © 2015 Wiley Periodicals, Inc.« less

Search for:
All Records
Creator / Author
"Zhang, Haoran"

Refine by:
Article Type
Availability
Journal
Creator / Author
Publication Date
Research Organization