Data-Driven Weather Forecasting and Climate Modeling from the Perspective of Development

Atmosphere, Год журнала: 2024, Номер 15(6), С. 689 - 689

Опубликована: Июнь 6, 2024

Accurate and rapid weather forecasting climate modeling are universal goals in human development. While Numerical Weather Prediction (NWP) remains the gold standard, it faces challenges like inherent atmospheric uncertainties computational costs, especially post-Moore era. With advent of deep learning, field has been revolutionized through data-driven models. This paper reviews key models significant developments modeling. It provides an overview these models, covering aspects such as dataset selection, model design, training process, acceleration, prediction effectiveness. Data-driven trained on reanalysis data can provide effective forecasts with accuracy (ACC) greater than 0.6 for up to 15 days at a spatial resolution 0.25°. These outperform or match most advanced NWP methods 90% variables, reducing forecast generation time from hours seconds. reliably simulate patterns decades 100 years, offering magnitude savings competitive performance. Despite their advantages, have limitations, including poor interpretability, evaluating uncertainty, conservative predictions extreme cases. Future research should focus larger integrating more physical constraints, enhancing evaluation methods.

Язык: Английский

---

Laterally gated ferroelectric field effect transistor (LG-FeFET) using α-In2Se3 for stacked in-memory computing array

Nature Communications, Год журнала: 2023, Номер 14(1)

Опубликована: Окт. 25, 2023

Abstract In-memory computing is an attractive alternative for handling data-intensive tasks as it employs parallel processing without the need data transfer. Nevertheless, necessitates a high-density memory array to effectively manage large volumes. Here, we present stacked ferroelectric comprised of laterally gated field-effect transistors (LG-FeFETs). The interlocking effect α-In 2 Se 3 utilized regulate channel conductance. Our study examined distinctive characteristics LG-FeFET, such notably wide window, effective switching, long retention time (over × 10 4 seconds), and high endurance 5 cycles). This device also well-suited implementing vertically structures because decreasing its height can help mitigate challenges associated with integration process. We devised 3D structure using LG-FeFET verified feasibility by performing multiply-accumulate (MAC) operations in two-tier configuration.

Язык: Английский

---

Review of Emerging Atomically Precise Composite Site‐Based Electrocatalysts

Advanced Energy Materials, Год журнала: 2023, Номер 13(37)

Опубликована: Авг. 6, 2023

Abstract Atomically precise composite site‐based catalysts with new electrocatalytic synergistic mechanisms and enhanced activities have emerged as a frontier in the electrocatalysis community. This topical review focuses on recent research advances of atomically metal sites‐based electrocatalysts. work first demonstrates an overview configurations sites, including discussion advanced methods employed for understanding sites. The then provides comprehensive organization previously reported methodologies synthesizing electrocatalysts Representative case studies are provided, starting from simple one‐step pyrolysis strategy to species‐by‐species multi‐step strategy. Based preceding discussions catalyst materials, further discusses unique raised by that different routine single species systems mainly involve oxygen reduction reaction, evolution hydrogen nitrogen carbon dioxide reaction. themes this section include true active center determination sites various types synergy mechanisms. Finally, critical unanswered questions remaining challenges, well promising underexplored directions identified.

Язык: Английский

---

Nonvolatile resistive switching memory behavior of the TiOx-based memristor

Chemical Physics, Год журнала: 2024, Номер 580, С. 112217 - 112217

Опубликована: Фев. 2, 2024

Язык: Английский

---

Precise Capture and Dynamic Release of Circulating Liver Cancer Cells with Dual‐Histidine‐Based Cell Imprinted Hydrogels

Advanced Materials, Год журнала: 2024, Номер 36(27)

Опубликована: Апрель 24, 2024

Abstract Circulating tumor cells (CTCs) detection presents significant advantages in diagnosing liver cancer due to its noninvasiveness, real‐time monitoring, and dynamic tracking. However, the clinical application of CTCs‐based diagnosis is largely limited by challenges capturing low‐abundance CTCs within a complex blood environment while ensuring them alive. Here, an ultrastrong ligand, l ‐histidine– ‐histidine (HH), specifically targeting sialylated glycans on surface CTCs, designed. Furthermore, HH integrated into cell‐imprinted polymer, constructing hydrogel with precise imprinting, high elasticity, satisfactory compatibility, robust anti‐interference capacities. These features endow excellent capture efficiency (>95%) for peripheral blood, as well ability release controllably Clinical tests substantiate accurate differentiation between cancer, cirrhosis, healthy groups using this method. The remarkable diagnostic accuracy (94%), lossless material reversibility, cost‐effectiveness ($6.68 per sample) make HH‐based potentially revolutionary technology single‐cell analysis.

Язык: Английский

---

The recent advances in the approach of artificial intelligence (AI) towards drug discovery

Frontiers in Chemistry, Год журнала: 2024, Номер 12

Опубликована: Май 31, 2024

Artificial intelligence (AI) has recently emerged as a unique developmental influence that is playing an important role in the development of medicine. The AI medium showing potential unprecedented advancements truth and efficiency. intersection to revolutionize drug discovery. However, also limitations experts should be aware these data access ethical issues. use techniques for discovery applications increased considerably over past few years, including combinatorial QSAR QSPR, virtual screening,

Язык: Английский

---

Boron-induced microstructural manipulation of titanium and titanium alloys in additive manufacturing

Virtual and Physical Prototyping, Год журнала: 2023, Номер 18(1)

Опубликована: Авг. 30, 2023

While the role of boron (B) has been thoroughly clarified in titanium (Ti) castings, microstructural changes triggered additive manufacturing (AM) are still subject debate literature. Many contributions have confirmed B-induced refinement Ti-based AM parts. The formation TiB matrix composites (TMCs) may increase strength. In some cases, B also promote columnar-to-equiaxed transition, thus mitigating anisotropic effects associated with strong epitaxial growth unidirectional columnar grains typical AM. However, as critically discussed this review, pitfalls remain. Due to fast cooling, evolution deviate from equilibrium, leading a shift Ti-B eutectic point and out-of-equilibrium phases. Additionally, undermine ductility crack propagation resistance parts, which calls for appropriate remediation strategies.

Язык: Английский

---

Analysis of Particle Size in Composite Materials Using Image Processing

International Journal of Engineering, Год журнала: 2024, Номер 37(4), С. 579 - 587

Опубликована: Янв. 1, 2024

Composite materials are the most important in science and engineering, which contain two or more materials. In scanning electron microscopy (SEM) technique is an approach to measure material''s particle size. A new procedure was used instead of SEM called Artificial Intelligence (AI). (AI) interdisciplinary branch computer that involves solving problems require human intelligence capabilities. The vision a subfield AI, uses some algorithms detect details images by using image processing. Detecting particles measuring size scanned essential task helps describe their feature, traditionally, calculated manually adding mesh drawing diagonal line arbitrary particle. this paper, model based on proposed analyze all particles. This additives composite like graphene flakes them depending reference fixed microscope (SEM). Open-source Computer Vision (OpenCV) library, utilizing multi-layers canny edge detection, Sobel filter, Brightness contrast algorithms, Python 3. results have achieved very satisfied indication with low process time = 0.2 mili-seconds.

Язык: Английский

---

A photonics perspective on computing with physical substrates

Reviews in Physics, Год журнала: 2024, Номер 12, С. 100093 - 100093

Опубликована: Июнь 15, 2024

We provide a perspective on the fundamental relationship between physics and computation, exploring conditions under which physical system can be harnessed for computation practical means to achieve this. Unlike traditional digital computers that impose discrete nature continuous substrates, unconventional computing embraces inherent properties of systems. Exploring simultaneously intricacies implementations applied computational paradigms, we discuss interdisciplinary developments computing. Here, focus potential photonic substrates computing, implementing artificial neural networks solve data-driven machine learning tasks. Several network are discussed, highlighting their advantages over electronic counterparts in terms speed energy efficiency. Finally, address challenges achieving programmability within outlining key strategies future research.

Язык: Английский

---

From Experimental Values to Predictive Models: Machine Learning‐Driven Energy Level Determination in Organic Semiconductors

Advanced Energy Materials, Год журнала: 2025, Номер unknown

Опубликована: Янв. 8, 2025

Abstract The precise determination of ionization energy (IE) and electron affinity (EA) is crucial for the development optimization organic semiconductors (OSCs). These parameters directly impact performance electronic devices. Experimental techniques to measure IE EA, such as UV photoelectron spectroscopy (UPS) low‐energy inverse (LE‐IPES), are accurate but resource‐intensive limited by their availability. Computational approaches, while beneficial, often rely on gas‐phase calculations that fail capture solid‐state phenomena, leading discrepancies in practical applications. In this work, machine learning methods used develop a chained model estimating EA values. By implementing transfer strategy, challenge experimental data effectively addressed, utilizing large database intermediate properties enhance training. efficacy demonstrated through its achieving mean absolute errors 0.13 0.14 eV respectively. has also been tested an external validation dataset comprising newly measured molecules. findings highlight potential OSC research, significantly enhancing property accessibility accelerating molecular design discovery.

Язык: Английский

---