32584 research outputs found
Sort by
Impervious Surfaces from High Resolution Aerial Imagery: Cities in Fresno County
Get PDFThis study investigates impervious surfaces — areas covered by materials with restricted water permeability, such as pavement, sidewalks, and parking lots—due to their crucial role in influencing water dynamics within urban landscapes. The impermeability of these surfaces disrupts natural water absorption processes, resulting in adverse environmental consequences such as increased flooding, erosion, and water pollution. The research employs impervious surface analysis, a method involving the mapping and analysis of these surfaces within specified study areas, including cities, counties, and census tracts. Remote sensing techniques, specifically satellites and aerial imagery, are commonly utilized for the identification and classification of impervious surfaces. In the context of Fresno County, diverse classification methods, encompassing pixel-based, object-based, and deep learning approaches, are employed to classify and evaluate impervious surfaces. Significantly, the deep learning classification method exhibits exceptional performance, achieving an impressive overall accuracy ranging between 85-92%. The study reveals that the estimated percentage of impervious surfaces in Fresno County cities approximates 45%, comparable to the characteristics of medium density residential areas. Noteworthy is the observation in the Fresno/Clovis city area, where the percentage of impervious surfaces escalated from 53% in 2010 (per EnviroAtlas) to 63% in 2020. This 10% increase over a decade closely aligns with concurrent population growth trends in the region. In conclusion, this research underscores the critical significance of comprehending and monitoring impervious surfaces due to their pivotal role in shaping the environmental quality and resilience of urban areas. The insights gleaned from this study provide valuable guidance for the development of effective land use planning and management strategies, specifically tailored to mitigate the adverse impacts of impervious surfaces on the environment and human well-being
Exploring Performance-Based Contracts: A Good Option to Address Long-Term Road Maintenance in California?
Get PDFPerformance-based contracts (PBCs) used in road maintenance provide flexibility for considering new materials, design, and technology to achieve predetermined performance targets. Several states in the U.S. have already used these types of contracts in road maintenance, and their experience can inform the use of PBCs in California. The objectives of this research are twofold. First, identify the benefits and challenges of PBCs compared to traditional contracts. Second, explore the main aspects of PBCs implementation for road maintenance and relate them to California\u27s context. To this end, deductive-inductive content analysis was conducted on 84 peer-reviewed articles published between 1998 and 2023. Findings of this research identified the main benefits of PBCs to be: (1) cost savings, (2) improved work/service quality, and (3) reduction in risk to the transportation agency through the transfer of responsibility to the contractor. The main challenges were found to be: (1) the need for training and a shift in mindset from traditional contracting forms to PBCs, (2) the need to establish trust between contracting agencies and contractors, and (3) the temptation for contractors to abuse the system. Research findings highlighted procurement, performance indicators, and incentives/disincentives as the three major themes to consider in PBC implementation. California does not have a history of using performance-based contracts for road maintenance. Results from this research may be the first stepping stone to initiate the decision-making process to use these types of contracts for road maintenance
Reinforcement Learning-based Dynamic Pricing for Revenue Maximization with Elastic Network Slicing
Get PDFNetwork slicing is a key enabler of next-generation networking that supports a diverse array of network applications with different service requirements. In particular, elastic network slicing that dynamically scales the bandwidth reserved for each network slice would benefit both slice users and providers through cost-effective resource utilization. However, the elasticity poses a complex problem of managing the dynamics of fluctuating network slices. It is necessary for a slice provider to maintain the balance of different types of slice requests, so it can accommodate more requests while satisfying the service requirements for each slice type. Dynamic pricing of slice resources is a way for a network operator to realize the ideal balance of different types of network slices by implicitly communicating the current network state to slice users. In this project, we formulate an online pricing scheme for elastic network slices, which maximizes the revenue of slice providers. Our problem considers (1) slice users’ a priori preference over different types of network slices (susceptibility to Service Level Agreement violations) and (2) the influence of prices on slice users’ decision to choose a type of slice services. Our simulation experiments in a practical network topology demonstrate the revenue increase of a network operator by encouraging the use of elastic network slices
Echo: A browser extension that runs experimental JavaScript
Get PDFNarcissus is a JavaScript interpreter written in JavaScript. While it is a good engine for experimenting with JavaScript’s design, it does not integrate easily into the browser. This project introduces ‘‘Echo’’, a browser add-on designed to execute Narcissus JavaScript files and scripts within web browsers. The project explores the performance of the Narcissus interpreter against native browser JavaScript engines and benchmarks the results, showcasing the trade-offs in running an experimental engine—the Narcissus interpreter—on the browser versus native JavaScript. Additionally, as a proof of concept, we implement taint tracking, a capability meant to boost security by preventing sensitive data from being corrupted. This builds on prior uses of Narcissus for security and language feature prototyping
Scalable Container Caching Optimization with Action Masking for Serverless Edge Computing
Get PDFServerless edge computing is an emerging technology that realizes the low latency and resource-efficient function calls for responsive computing. In cloud-based serverless computing, it is a common practice to cache sufficiently many function containers for future reuse to reduce the overhead of container initiation. In contrast, the capacity limitation of edge nodes poses a complex problem to the caching strategy in serverless edge computing of selecting an appropriate set of container caches based on the request distribution. Deep Reinforcement Learning (DRL) can play a crucial role in optimizing the caching decisions under dynamic request arrivals. In this paper, we propose a scalable container caching agent based on DRL to improve the training efficiency and caching performance. We propose to apply action masking, which eliminates inauspicious caching actions from the DRL exploration and concentrates computing resources to more promising actions, to achieve scalability. Our proposal is evaluated through comparative analysis by simulations based on the metrics of latency, training time, and cache hit rate. Our results suggest that DRL with action masking can efficiently solve the combinatorial optimization of container caching and further infer the capability of DRL in solving similar combinatorial optimization problems in networking
Fake Malware Generation using GANs as API Calls
Get PDFWith malware threats on the rise, they have also grown more complicated and subtle. Consequently, incorporating cutting-edge machine learning into cybersecurity defenses has never been more crucial. Nevertheless, building resilient machine-learning models is a significant challenge due to the need for existing diversified and complete malware datasets. This project will relieve this difficulty by employing a Generative
Adversarial Network (GAN) to develop artificial malware samples featuring an Appli- cation Programming Interface (API) call series. While traditional generative modeling
has primarily been limited to image-based fields, we offer an “outside the box” domain – malware signature generation – as an API call sequence. The goal is to imagine synthetic malware that highly resembles benign malware; several GAN architectures specializing in sequence generation will be employed to generate “fake” malware that appears authentic. Researchers can use synthetic datasets to exploit the scarcity of data for training machine learning, but they also help detection models bridge the gap and mimic this “real” malware. Preliminary results show our approach is on a favorable course, with synthetic samples exhibiting enough lifelike traits that current models do not accurately detect, creating a foundation for more robust and adaptable malware detection systems
NFT Price Prediction using Machine Learning
Get PDFIn the evolving cryptocurrency marketplace, Non Fungible Tokens (NFT) pose a unique challenge when it comes to predicting the prices due to their high volatility and fluctuating nature. This project aims to create a model that utilizes deep learning techniques to accurately forecast NFT prices. Based on the real time data and the transaction history, the model uses Convolutional Neural Networks (CNNs) and Long Term Short Memory Networks (LSTM) to analyze and make effective predictions about future prices. The methodology involves gathering data from two online marketplaces, Dune and Opensea to create a dataset that enhances the model’s predictive capabilities. This research includes data collection, data preprocessing, and model evaluation ensuring a validation process for the proposed predictive system. The primary objective of the project is to provide analytics to the investors and the analysts for making informed decisions, within the NFT market
Deception Detection Models from Speech
Get PDFRecently, researchers have shown an increased interest in automatically detecting deceptive actions. The attention given to this area can be attributed to the many potential applications of deception detection, especially in the field of criminology. To contribute to the deception detection research, this project investigates textual and audio data extracted from spoken and written words. We evaluated and compared the traditional linguistic models with advanced Large Language Models (LLMs) while using Natural Language Processing (NLP) techniques. Additionally, various feature selection techniques were applied to assess the importance of linguistic features. We conducted extensive experiments to evaluate the effectiveness of both conventional and deep NLP models on textual data. In addition, the deep models were also applied to audio data. Findings suggest that the best-performing models for each data type are the Bidirectional Long Short Term Memory for textual data and the ResNet50 for audio data. These models were then combined to create a late fusion model that outperforms other text and audio models from previous research using the measures of accuracy and F1 score for comparison. This late fusion model achieved an impressive score with an accuracy of 90.9% and an F1 score of 91.07%
PROTOCOL: Street outreach conflict mediation programs for reducing violence
Get PDFThis is a protocol for a Cochrane Review (intervention). The objectives are as follows: This systematic review will synthesize the available evaluation research on the effectiveness of street outreach conflict mediation programs. The review seeks to answer the following primary question: Are street outreach worker strategies that use conflict mediation and/or violence interruption strategies effective at reducing violence? Assuming a sufficient number of eligible studies, this review will also address two additional questions: Are there certain program elements that render these strategies more or less effective? Are there certain conditions under which these strategies are more or less effective? As policymakers struggle to understand the policy options available to them for preventing and reducing violence, having clear answers to these three questions will help them make more informed decisions. The primary focus of this review is the effect of these strategies on violence. Nonetheless, when data are available we will collect information on secondary outcomes such as the cost-effectiveness of these strategies and their impacts on perceptual or attitudinal measures such as fear, perceived safety, and violence-related norms