教学课程资源库 更多>>

关键词： Metadata   Computer science   Deep learning   Tools   Support vector machines   Licenses   Libraries   Research paper classification   Word2Vector (W2V)   metadata   association of computing machinery (ACM)   k-nearest neighbor's (KNN   decision tree (DT)   random forest (RF)   term frequency (TF)   term frequency and inverse document frequency (TFIDF)   bag of word (BOW)  

摘要： The existing plethora of document classification techniques exploits different data sources either from the content or metadata of research articles. Various journal publishers like Springer, Elsevier, IEEE, etc., do not provide open access to the content of research articles, whereas metadata is freely available there. Metadata like title, keyword, and abstract can serve as a better alternative to the content in various scenarios. In the current literature, researchers have assessed the role of some of the metadata individually. We believe that the collective contribution of metadata parameters can play a significant role in classifying research papers. This paper presents a comprehensive evaluation of the role of metadata, individually as well as in combinations to achieve the objective of research paper classification. Moreover, we have classified the research articles into ACM hierarchy root categories (e.g. general literature, hardware, software, etc.). In this comprehensive evaluation, we have assessed all the possible combinations of metadata features against different classifiers such as Random Forest, K Nearest Neighbor, and Decision Tree. The results of this research reveal that the title & keywords combination outperforms other combinations with an F-measure score of 0.88.

关键词： Computer science   Artificial intelligence  

摘要： Cost-effective approaches are a major concern for water facilities in general. Thus, resource recovery in wastewater treatment process has been innovated within water resource recovery facilities (WRRF) and achieved by pretreatment technologies within the anaerobic digestion (AD) such as thermal hydrolysis pretreatment (THP); results are reduced cost and advance the treatment process (Barber, 2016). This thesis focuses on an acoustic based approach for the quantification of the physical state of the THP sludge known as gelation factor. Recent work has shown that the gelation behavior of sludge is linked to decreased hydrolysis rates and biodegradability (Rasel et al., 2019). Thus, gelation factor is an important parameter for optimization of THP conditions and advocated directions for full-scale facilities to maximize the benefit of THP process. A method combining the low maintenance and data rich acoustic chemometric as a process analytical technology with multivariate machine learning algorithm to predict the gelation factor were developed within this study to enhanced process stability, capacity of current systems and optimized designs for future systems.

关键词： Nutrition   Design   Physiology   Aging   Biology   Computer science   Fluid mechanics   Gerontology   Mathematics   Mechanics   Medicine   Morphology   Pharmacology   Physics   Public health  

关键词： Computer science   Medical imaging   Artificial intelligence  

关键词： Computer science   Artificial intelligence  

摘要： In this work, we tackle the question: Can neural networks count? More precisely, given an input image with a certain number of objects, can a neural network tell how many are there? To study this, we create a synthetic dataset consisting of black and white images with variable numbers of white triangles on a black background, oriented right-side up, down, left or right. We train a network to count the right-side up triangles; specifically, we see this as a closed-set classification problem where the class is the number of right-side up triangles in the image. These evaluations show that our networks, even in their simplest designs, are able to count a particular object in an image with a very small epsilon of approximation. We conclude that the neural networks are enforced with more complex learning capabilities than given credit for.

关键词： Energy   Chemistry   Aerospace engineering   Analytical chemistry   Computer science   Engineering   Materials science   Medical imaging   Physical geography  

摘要： Graphitic materials describe those containing carbon within a honeycomb lattice structure and have attracted vast interest due to their novel and advantageous properties. Whilst large amounts of work have been conducted to investigate the nature of such materials, in the context of graphite, graphene and other carbon allotropes, there are still many voids in our understanding of these materials, extending to their structure, modification, characterisation, and application. Plasma-exfoliation offers a facile approach towards the large-scale synthesis of partially oxidised, multilayered stacks of graphitic layers. Within the scope of this thesis, plasma-exfoliated multilayer graphitic materials (NPs) are investigated in detail using a vast array of characterisation techniques to reveal information regarding their morphology, structure and elemental composition; including X-ray Photoelectron Spectroscopy (XPS), Raman spectroscopy, X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM), Atomic Force Microscopy (AFM), gas physisorption analysis, Nuclear Magnetic Resonance Spectroscopy (NMR), Fourier-Transform Infrared Spectroscopy (FT-IR), Thermogravimetric Analysis (TGA) and Energy-Dispersive X-ray (EDX) analysis. Various novel functionalisation strategies are also explored to covalently and non-covalently functionalise NPs. Initially, an aryl boronic acid precursor is utilised as a radical source to provide a facile route towards the covalent functionalisation of NP. Such a methodology serves to identify safer aryl radical precursors as an alternative to commonly employed diazonium salt reagents, which are typically hazardous and undesirable for large-scale application. This functionalisation strategy enables the covalent attachment of 4-(trifluoromethyl)phenyl moieties to the edges of NP1 stacks. A percentage atomic composition (at.%) of 3.5 fluorine incorporation is revealed via XPS and Raman spectroscopy, TGA, FT-IR and E

关键词： Software   Sensors   Neural networks   Artificial intelligence   Computer science  

摘要： During the last decade, automotive manufacturers have introduced increasingly capable driving automation functions in consumer vehicles. As the functionality becomes more advanced, the task of driving moves from the human to the car. Hence, making sure that autonomous driving (AD) functions are reliable and safe is of high importance. Often, increased levels of automation result in more complex safety validation procedures, that may be both expensive, time consuming, and dangerous to perform. One way to address these problems is to move parts of the validation to the virtual domain. In this thesis, we investigate methods for validating AD functionality in virtual simulation environments, using methods from machine learning and statistics. The main focus is on how to make virtual simulations resemble real-world conditions as closely as possible. We tackle this with an approach based on sensor error modeling. Specifically, we develop a statistical sensor error model that can be used to make ideal object measurements from simulations resemble measurements obtained from the perception system of a real-world vehicle. The model, which is based on autoregressive recurrent mixture density networks, was trained on sensor error data collected on European roads. The second part considers system falsification using reinforcement learning (RL); a flexible framework for validation of system safety, which naturally allows for the integration of, e.g., sensor error models. We compare results of system falsification using RL to an exact approach based on reachability analysis. With this thesis, we take steps towards more realistic statistical sensor error models for virtual simulation environments. We also demonstrate that approximate methods based on reinforcement learning may serve as an alternative to reachability analysis for validation of high-dimensional systems. Finally, we connect the RL falsification application to sensor error modeling as a possible direction for future re

关键词： Viruses   Viral infections   Coronaviruses   COVID-19   Virology   Computer science   Bioinformatics   Immunology   Pharmacology  

摘要： The human immune system is responsible for the detection and elimination of pathogens. The immune response to viral infection is comprised of a complex set of multi-scale interactions between small molecules, proteins, genes, and cells that govern pro-inflammatory and anti-inflammatory processes. Overactive inflammation is a root cause of severe clinical outcomes and can lead to systemic tissue damage and death. Identifying the drivers of dysregulation is key in the development and administration of therapeutic treatments to maintain equilibrium. The work presented aims to utilize high throughput and immunological data to determine the causal agents of immunoregulation using systems biology tools. Developed methods address this problem at the protein and systems levels using protein-protein interaction (PPI) network methods and ordinary differential equation (ODE) models. Aim 1, the creation of the first ever disease specific subnetwork, identifies a set of proteins that are enriched for possible antiviral drug targets for influenza A infection. In Aim 2, PPI network controllability analyses are used to identify a set of 24 and 16 proteins acting as regulators of influenza A and SARS-CoV-2 infection, respectively. These proteins are further prioritized as targets in drug development/repurposing based on topology, function, and known targeting compounds. Five previously unidentified compounds are recommended for repurposing to treat COVID-19. Together, Aims 1 and 2 computationally produce efficient and meaningful biological results which align with in vivo findings. Aim 3 explores influenza A strain-specific dynamics observed in immunological data by using ODE models to elucidate which biological mechanisms are at the root of differential behavior. Two models are constructed and parameterized to explore H1N1 and H5N1 influenza dynamics. Study reveals that only a small number of host functions likely contribute to the strain-specific response, particularly the product

摘要： Disability Inclusion and Accessibility in technology and education have today emerged as a necessity of national importance. National Science Foundation(NSF) funded projects like AccessCSforALL, Bootstrap, and CSforAll, are making efforts to make computer science inclusive to the 7.4 million K-12 students with disabilities. Bringing people with disabilities into tech workforce would not only create equal opportunities for all but would also better equip the society to serve its diverse population. Computer Programming is a growing and important field of the present and future. When the tools to learn programming from an early age are inaccessible to people with disabilities, they are denied the opportunity to succeed. This thesis focuses on making computer science concepts accessible to K-12 students who are Deaf/Hard of Hearing, Blind/Low Vision and those with motor disabilities. It attempts to provide technical strategies to break this inaccessibility barrier and make computer science literacy, education, and, ultimately, careers accessible to all.

关键词： Atmospheric sciences   Computer science   Artificial intelligence