10368 research outputs found
Sort by
Evaluating Corruption Defenses for Model Robustness
No full textWithin the last few years, deep learning models have seen increased usage in various computer vision domains and have attained high accuracy rates for several visual identification tasks. However, when faced with real data containing visible anomalies, their accuracy can be reduced significantly. In addition, adversarial examples, inputs with deliberate changes that induce misclassification, have further hindered their performance. The training methods used to minimize the effects of adversarial examples have also been found to negatively impact a model\u27s robustness on clean data as well as common noise and distortions, which may be an undesired trade-off. Adversarial robustness is a model\u27s ability to resist intentional deceptive inputs, while corruption robustness is its resilience to everyday noise and distortions. In our work, we adopt a technique that improves the neural network\u27s ability to generalize by training on augmented images of the dataset and evaluate its efficacy in the face of adversarial examples with several empirical metrics, comparing its performance with traditional adversarial training techniques
Using Machine Learning to Analyze and Predict Password Reset Factors
No full textLarge companies spend millions of dollars annually on password resets. While a lot of research has been done on the influence of password strength and memorability on resets, little public research has examined factors beyond just the password itself. In order to identify which factors most influence password resets, this study investigated device type, days since last login, and account age. The type of password reset was also categorized into four types: reset without trying, meaning that the user had not recently attempted a login before resetting; tried and failed, meaning that the user had attempted one login before resetting; and forced reset, meaning that the user had two or more failed login attempts, resulting in a subsequent password reset. Analyzing 200,000 rows of data revealed that newer accounts (8-31 days old) and those accessed from a mobile device were more likely to initiate password resets. It was also found that resetting without trying was the most common type of password reset in both desktop and mobile devices. Using this predictive model, organizations can preemptively address password reset issues, improving user experience and security while reducing company costs
The Future Of Biodegradable Technolgy Creation: Liquid Crystals
No full textNearly 400 million electronic items are dumped into landfills annually, and less than 20% of that e-waste is recycled due to their unsustainable compositions. This expels heavy metals that can damage our central nervous systems and water sources and releases a dangerous amount of greenhouse gasses into the environment. There is an urgent need for the creation of biodegradable technology. To help counter this issue, I employ a cellulose derivative, hydroxypropyl cellulose (HPC), and its potential as a dependable biodegradable material. HPC forms a liquid crystal phase with a chiral nematic structure. Due to their molecular properties, they are soluble in both water and organic solvents. They also have many attractive characteristics including angle-dependent optical properties and the ability to self-assemble, making them suitable candidates for programmable sensing technology creation. In my experiments, pure HPC solutions and solutions containing Carbon Black and Mxene as conductive additives were used to complete basic circuits to test their conductivity. Potentiostatic electrochemical impedance spectroscopy (EIS) was then used to quantify the conductive potential and resistivity for comparative analysis of the various solutions. The findings of this project are significant to the developmental field of biodegradable technology
They All: An Emerging Pronoun
No full textThis study explores the potential adoption of they all (or th’all ) as an emerging third-person plural pronoun that addresses linguistic and social issues posed by the singular they. While the singular they has been endorsed for its inclusivity, its dual usage in both the singular and plural cases can confuse readers and speakers, hindering its broader acceptance. Additionally, resistance to the singular they is often tied to deeply rooted societal views on gender and opposition to non-binary identities. Drawing inspiration from the evolution of the plural you into you all (in Southern American English), this research proposes they all as a parallel innovation to clarify plurality and reduce discomfort associated with pronoun use. A survey of Southern American English speakers was conducted to test this hypothesis, analyzing their perceptions of gendered language, political ideology, and familiarity with informal pronouns. Perceptions of gendered language refer to how participants rated selected words on how masculine or feminine they were, detailing their agreement with the concept of gender as a binary. Participants also evaluated audio clips containing various pronouns, including they all. This study hypothesizes that weaker adherence to the gender binary predicts higher acceptance of they all
Influence of Glucose Deprivation on Amyloid Beta Neurotoxicity
No full textAlzheimer’s disease (AD) is a neurodegenerative disorder that is identified by cognitive decline, including memory loss and an impaired ability to perform daily, normal tasks, which leads to significant emotional, physical, and financial burdens for patients and caregivers. The main cause of AD is the accumulation of amyloid beta (Aβ) plaques, which disrupt neuronal function and cause disease progression. Glucose, the brain\u27s primary energy source, plays a vital role in maintaining normal cognitive function. However, disruptions or issues in glucose metabolism are often developed in neurodegenerative diseases, including AD. This study explores the effects of glucose deprivation on amyloid beta neurotoxicity, and neuronal health in AD models, using SH-SY5Y neuroblastoma cells. The research investigates if glucose deprivation increases Aβ-induced neurotoxicity. SH-SY5Y cells were differentiated using retinoic acid (RA) and Phorbol-12-myristate-13-acetate (PMA) and exposed to varying glucose conditions along with varying concentrations of Aβ treatment. Cell viability and cytotoxicity were assessed using the Enzo LDH Assay, and the impact of glucose deprivation on Aβ cell death was measured. Though cytotoxic effects of Aβ on SH-SY5Y could be confirmed, results suggest that reduced glucose availability does not affect the intensity of neurodegenerative effects of Aβ. No statistically significant difference was observed in LDH release for the varying glucose concentrations in the presence of amyloid beta at 500 nM or 2 μM. This study aimed to enhance understanding of the connection between glucose metabolism and amyloid beta toxicity in Alzheimer\u27s. A negative effect of glucose deprivation on Aβ cytotoxicity could not be established
Exploring Methods to Decrease Congestion Within The Beijing–Hong Kong and Macau Expressway by Re-Working the Roadway Architecture Using AnyLogic 8.9
No full textTraffic congestion is an ever-growing problem that worsens as the population increases. This issue is particularly evident in populated countries like China, where traffic infrastructure is crucial when transporting goods and people. While there have been previous studies focusing on alleviating the traffic congestion on some Chinese roadways, there is a pressing need for studies specifically aimed at reconstructing the Beijing-Hong Kong and Macau Expressway. This research aims to design an efficient Beijing-Hong Kong and Macau Expressway model that allows for a greater vehicle output than the original model. It was hypothesized that if an improved model of the Beijing-Hong Kong and Macau Expressway was designed, then the vehicle output of the expressway would increase relative to the original model because of reduced congestion. Using AnyLogic software, three models were developed to represent the Beijing-Hong Kong and Macau Expressway: a control and two alternative models were designed (Alternative Model 1 & Alternative Model 2). The alternative models were designed based on previous studies aimed at alleviating traffic congestion, incorporating features such as U-turn and ramp metering lanes. When comparing the results to the control model, data showed significant increases in vehicle output in each alternative model compared to the control. Comparative analysis shows that there was a 109.64% difference in mean between the Control and Alternative Model 1 and a 99.83% difference between the Control and Alternative Model 2
Testing the Accuracy of QWaterModel Against Eddy Covariance Evapotranspiration Data of Almond Orchards Located in California Using Landsat Imagery
No full textThe paper investigates the efficacy of the QWaterModel, a QGIS plugin designed for estimating evapotranspiration (ET) in almond orchards located in California\u27s water-scarce San Joaquin Valley. Given the anticipated 25% rise in global water requirements for agriculture by 2080 and the increasing drought pressures, the need for innovative water management practices is paramount (Nikolaou et al., 2020). The research aims to compare ET estimates from QWaterModel with eddy covariance ET measurements (ECET) across two almond orchards (WWF and OLA) to assess the model\u27s reliability in real-world agricultural applications. Utilizing Landsat imagery and ground-based ECET data, the study finds a strong correlation (r = 0.87, R² = 0.75) between QWaterModel estimations and actual ET measurements, indicating that the model effectively captures the dynamics of water usage in almond cultivation. The QWaterModel\u27s user-friendly interface and integration with satellite data highlight its potential as a practical tool for water management, especially for farmers lacking advanced technical skills. Nevertheless, variability in estimates suggests that the model may oversimplify complex ET processes affected by local conditions, urging the need for further calibration and validation against diverse data sources. The findings underscore the importance of accurate ET estimation for sustainable agricultural practices amidst growing water scarcity and point out ways to enhance QWaterModel\u27s applicability
Chemically-triggered dopant release from surface modified polypyrrole films
Get PDFPolypyrrole (PPy) is cationic in its conducting form, requiring a charge-balancing counterion, or dopant. The release of bioactive dopants, driven by the reduction of PPy films, offers a route to controlled drug delivery. Thiol-terminated long chain poly (ethylene glycol) (PEG) reacts with a dodecylbenzene sulfonate (DBSA)-doped PPy, forming a dense overlayer and partially liberating DBSA via the chemical reduction of the film. The resulting PEG brush acts as a barrier to dopant diffusion from the film, but proteins have been shown to disrupt this layer, releasing the DBSA. The mechanism by which this disruption occurs has not been thoroughly investigated. In this study, dopant release from PEG-PPy composites was examined via systematic exposure to a variety of chemical stimuli, including macromolecules such as poly (ethylene imine), polyethylene glycol, and poloxamers, as well as small-molecular-weight alcohols, carboxylic acids, and amines. Dopant release was quantified by quartz crystal microbalance. Poly (ethylene imine) efficiently released DBSA, while anionic and uncharged macromolecules did not. All classes of small molecules triggered dopant release, with longer homologues magnifying the response. The mechanisms of dopant removal are dependent on the functional groups of the stimulating agent and include ion exchange and nucleophilic reduction of the polycationic backbone. Tosylate, salicylate, and penicillin dopants showed release behaviors similar to DBSA, demonstrating the generality of the PEG barrier
2025 Sub-Librarians Meeting: The Adventure of Lomax the Sub-librarian Featuring Lyndsay Faye
No full textALA Best Historical Award-winning (and twice Edgar Award-nominated) Sherlockian author, Lyndsay Faye, will give a presentation on Lomax--the sub-librarian of the British Library and namesake of the Sherlock Holmes society, the Sub-librarians Scion of the Baker Street Irregulars. Her comments will draw upon her own research and her experience in writing her short story “The Gospel of Sheba,” a bibliomystery centered around a lethal grimoire, narrated via Lomax\u27s private journals. The presentation will be followed by 15 minutes of Q&A
