background introduction by Chingfen Chang

Attention Is All You Need

2025-03-20 03:22:24 UTC

Recurrent neural networks, long short-term memory [12] and gated recurrent [7] neural networks in particular, have been firmly established as state of the art approaches in sequence modeling and transduction problems such as language modeling and machine translation [29, 2, 5]. Numerous efforts have since continued to push the boundaries of recurrent language models and encoder-decoder architectures [31, 21, 13].

Integrated multimodal artificial intelligence framework for healthcare applications

2025-03-20 03:22:32 UTC

Artificial intelligence (AI) and machine learning (ML) systems are poised to become fundamental tools in next-generation clinical practice and healthcare operations¹. Such anticipated utility, particularly in AI/ML systems aimed to improve clinical efficiency and patient outcomes, will require knowledge from multiple data sources and various input modalities²^,³^,⁴. Multimodal architectures for AI/ML systems are attractive because they can emulate the input conditions that clinicians and healthcare administrators currently use to perform predictions and respond to their complex decision-making landscape²^,⁵. A typical clinical practice uses a diverse set of information formats contained within the patient electronic health record (EHR) such as tabular data (e.g., age, demographics, procedures, history, billing codes), image data (e.g., photographs, x-rays, computerized-tomography scans, magnetic resonance imaging, pathology slides), time-series data (e.g., intermittent pulse oximetry, blood chemistry, respiratory analysis, electrocardiograms, ultra-sounds, in-vitro tests, wearable sensors), structured sequence data (e.g., genomics, proteomics, metabolomics) and unstructured sequence data (e.g., notes, forms, written reports, voice recordings, video) among other sources⁶.

2025-03-20 03:23:18 UTC

MICRO-ELECTRO-MECHANICAL Systems (MEMS) resonator is a versatile electromechanical component that could be used for sensing [1] and analog signal processing [2] in smart sensor systems.

OPEN-WORLD REINFORCEMENT LEARNING OVER LONG SHORT-TERM IMAGINATION

2025-03-20 03:23:21 UTC

Open-world decision-making in the context of reinforcement learning (RL) involves the following characteristics: (i) The agent operates within an interactive environment that features a vast state space; (ii) The learned policy presents a high degree of flexibility, allowing interaction with various objects in the environment; (iii) The agent lacks full visibility of the internal states and physical dynamics of the external world, meaning that its perception of the environment (e.g., raw images) carries substantial uncertainty.

2025-03-20 03:23:48 UTC

Ferroelectric materials exhibit stable states of collectively ordered electrical dipoles whose polarization can be reversed under an applied electric field5. Consequently, ultrathin ferroelectrics are of great technological interest for high-density electronics, particularly field-effect transistors and non-volatile memories2.

2025-03-20 03:23:54 UTC

As energy technology advances rapidly, the importance of rational

energy planning escalates. The pursuit of sustainable energy use has emerged as a focal point in scientific research [1,2]

Among the numerous ways to address 3D stacking of

2025-03-20 03:24:12 UTC

integrated circuits (e.g. through-silicon via, micro copper

pillar), a promising method is Cu/SiO2 hybrid bonding.

This approach enables to achieve mechanical and

electrical connections all at once and the interconnection

pitch is currently limited only by alignment accuracy [1].

Furthermore, hybrid bonding is compatible with standard

CMOS processes and mechanically strong, making it a

cost-effective and reliable solution compared to other

wafer bonding and assembly schemes for 3D integration

(such as metal/adhesive bonding, wire-bonding, etc.).

Transition metal dichalcogenides (TMDs) have demonstrated immense potential for applications in state-of-the-art electronic, optoelectronic, and spintronic devices because of their outstanding electronic, optical, mechanical, and magnetic properties.

2025-03-20 03:24:23 UTC

IB

2025-03-20 03:24:25 UTC

TRANSISTOR models are important to the semiconductor industry, which need fast and accurate models for circuit simulation and design optimization. Industry-standard compact models, such as the Berkeley short-channel IGFET model (BSIM) series of models [1], [2], use physics-based equations.

2025-03-20 03:24:38 UTC

Owing to the atomic-scale thickness

The surface free of lateral dangling bonds of 2DSCs limits the amount of trap states, and their atomic-scale flatness reduces the scattering from roughness

Monolayer (ML) WS2 is a member of TMDs with a finite band gap reported in a range of 1.8 ∼ 2.1 eV.[17] ML WS2 has a reasonably high carrier mobility of about 214/486 cm2 V−1s−1 at room temperature/5 K

2025-03-20 03:24:43 UTC

Whereas pure 0-1 and integer programming problems are favorite topics for theoretical research and experimentation in operations research, mixed integer programming has received less attention. This situation is unfortunate, since practitioners seem to generally agree that most applications involving integer variables are mixed integer programs, and often mixed 0-1 programs.

YOLOv4: Optimal Speed and Accuracy of Object Detection

2025-03-20 03:24:48 UTC

The majority of CNN-based object detectors are largely applicable only for recommendation systems. For example, searching for free parking spaces via urban video cameras is executed by slow accurate models, whereas car collision warning is related to fast inaccurate models. Improving the real-time object detector accuracy enables using them not only for hint generating recommendation systems, but also for stand-alone process management and human input reduction

2025-03-20 03:24:52 UTC

Graphics Processing Units (GPUs) have become important in providing processing power for high performance computing applications. CUDA [7] and Open Computing Language (OpenCL) [11] are two interfaces for GPU computing, both presenting similar features but through different programming interfaces. CUDA is a proprietary API and set of language extensions that works only on NVIDIA’s GPUs. OpenCL, by the Khronos Group, is an open standard for parallel programming using Central Processing Units (CPUs), GPUs, Digital Signal Processors (DSPs), and other types of processors.

Background (IB)

2025-03-20 03:24:55 UTC

Nowadays, photodetectors (PDs), used to convert incident optical signals into electrical signals,¹ play a crucial role in human life due to their applications in many fields, such as optical communication, security, etc.² In recent decades, various investigations into different materials have been conducted with the aim of decreasing the cost and miniaturizing the device size. Among them, metal oxide semiconductors (MOSs) are considered highly suitable candidates for photodetectors owing to their promising attributes, such as excellent electronic properties, wide bandgap energy, etc.

2025-03-20 03:25:24 UTC

The multi-item capacitated lot sizing problem (MICLS) belongs to medium-term production planning decisions and aims at determining the quantity (or lot sizes) and timing of production for several products over a number of finite production periods so as to satisfy a known demand for each product in each period, while respecting a known capacity in each period and minimizing the overall costs involved. The output of this problem is a production plan with the quantities of each product to produce in each period and the quantities of each product to store at the end of each period, so as to satisfy future demands. Frequently, in production planning systems this problem is called master production schedule.

The MICLS is a multi-item problem, with limited capacities and with setup times, has a single level because it only considers final products, has deterministic and dynamic demands, and is NPhard.

A human-airway-on-a-chip for the rapid identification of candidate antiviral therapeutics and prophylactics

2025-03-20 03:26:30 UTC

The increasing incidence of potential pandemic viruses—such as influenza A virus, Middle East respiratory syndrome coronavirus, severe acute respiratory syndrome coronavirus (SARS-CoV) and now SARS-CoV-2—requires development of new preclinical approaches that can accelerate the development of effective therapeutics and prophylactics. One of the most rapid ways to confront a pandemic challenge would be to repurpose existing drugs that are approved for other medical indications as antiviral therapeutics or prevention therapies. Although researchers and clinicians around the world are attempting to do this for the coronavirus disease 2019 (COVID-19) pandemic, the current approaches have been haphazard and generally rely entirely on the results of in vitro screens with cell lines.

Educators have emphasized the importance of fostering students’ ability to apply interdisciplinary knowledge for problem solving (Ivanitskaya, Clark, Montgomery, & Primeau, 2002). Integrative activities refer to the learning tasks engaging students in developing projects or solving problems using the knowledge and skills learned from multiple courses (Bannan-Ritland, 2003).

2025-03-20 03:26:33 UTC

2025-03-20 03:29:15 UTC

Ferrofluids are unique and versatile colloidal suspensions composed of nanoscale magnetic particles, typically magnetite, dispersed in a carrier fluid such as oil or water. These particles, approximately 10 nanometers in diameter, are coated with a surfactant to prevent agglomeration and maintain stability.