Foothill College

Cloud computing has become crucial for the commercial world due to its computational capacity, storage capabilities, scalability, software integration, and billing convenience. Initially, clouds were relatively homogeneous, but now diverse machine configurations in heterogeneous clouds are recognized for their improved application performance and energy efficiency. This shift is driven by the integration of various hardware to accommodate diverse user applications. However, alongside these advancements, security threats like micro-architectural attacks are increasing concerns for cloud providers and users. Studies like Repttack and Cloak & Co-locate highlight the vulnerability of heterogeneous clouds to co-location attacks, where attacker and victim instances are placed together. The ease of these attacks isn’t solely linked to heterogeneity but also correlates with how heterogeneous the target systems are. Despite this, no numerical metrics exist to quantify cloud heterogeneity. This paper introduces the Heterogeneity Score (HeteroScore) to evaluate server setups and instances. HeteroScore significantly correlates with co-location attack security. The paper also proposes strategies to balance diversity and security. This study pioneers the quantitative analysis connecting cloud heterogeneity and infrastructure security.

The UN reports that almost 700 million people can't afford food, while 1.3 billion tons of food is wasted yearly. The waste happens throughout the process and many factors contribute to it. The solution presented in this study utilizes a multimodal deep learning network to predict the decay of agricultural produce several days in advance and enables timely interventions to reduce spoilage and enhance food accessibility and affordability. A new set of data was collected to analyze the decay pattern of the selected agricultural produce. Raw tomatoes and strawberries were used for the study and were placed at multiple locations to capture their decay patterns. Temperature, humidity, and images were collected at regular intervals until the produce was considered decayed. This dataset trained detection AI models to isolate produce in large batches, and multimodal regression networks to forecast decay. A model that provided the least error in decay prediction was selected by experimentation of different models. The selected models were then deployed to a Raspberry Pi system with a camera, temperature sensors, and humidity sensors to prototype how well the trained model would perform in field deployment scenarios.

Humans and chess engines exhibit significant disparities in play style, prompting many top players to adopt strategies similar to those of chess engines. However, when playing against humans, several confounding factors, such as how the players may not be particularly dependent solely on skill, but also psychology, and visible emotions, influence the game beyond mere skill. While chess engines consistently play the best moves, humans often rely on intuitive and practical moves that can create counter-play. Given the constraints and complexity of the board, it is more effective to teach beginners through a multi-month workshop focused on studying and learning about positional and human aspects of chess, exploiting the opponent's psychological state and time pressure. We conducted a study with 50 novice chess players, grouped by age, to test the effectiveness of these differing teaching methodologies. Our analysis indicates a significant increase in improvement with teaching using a human teaching methodology in comparison to teaching using a chess engine teaching methodology.

Finding the method to most efficiently allocate resources in cloud computing environments has long been a challenge facing the cloud computing field, necessitating unique strategies to optimize performance while minimizing costs. The dynamic nature of cloud computing environments paired with users’ desire for cost-effective solutions requires intelligent and adaptive resource allocation methods. In this paper, we investigate the possibility of utilizing reinforcement learning (RL) algorithms to effectively handle this resource allocation problem. The findings provide insight into the possibility of using RL algorithms for cloud resource management and provide a base for further research and exploration in the area.

Maternal smoking during pregnancy is a significant public health concern that negatively affects fetal development, leading to complications such as low birth weight and increased risk of sudden infant death syndrome (SIDS) and affecting crucial biological processes. Cord blood cotinine, a biomarker of nicotine exposure, serves as a reliable indicator of fetal exposure to tobacco smoke, including secondhand and thirdhand smoke. My research develops a machine learning framework that combines clinical and genetic markers to predict neonatal cord blood cotinine levels, a biomarker of tobacco smoke exposure, in the absence of prior information about maternal smoking. The framework utilizes a dataset comprised of clinical and genomic information from cord blood, maternal blood, and placental samples, sourced from the National Center for Biotechnology Information (NCBI) platform. Four machine learning algorithms-K-Nearest Neighbor, Multi-Layer Perceptron, XGBoost, and Random Forest-were employed in the framework’s training and testing phases. XGBoost emerged as the most effective algorithm, achieving the best overall performance across four evaluation metrics (RMSE $=20.92$, MAE $=8.79$, MAPE $\left.=9.89, R^{2}=0.31\right)$. The results underscore the potential of this methodology in accurately predicting cord blood cotinine levels, thereby facilitating the identification of tobacco exposure in neonates and enabling the implementation of targeted health interventions to mitigate associated risks.

The overseas expansion of Chinese brands and the presentation of their video advertisements has been a crucial topic in media studies. Researchers have found that more and more Chinese companies are expanding abroad in the context of economic globalization, yet there is still a research gap on the issue of advertising in overseas markets. Therefore, this paper aims to further analyses the similarities and differences in advertising presentation between two Chinese local cosmetic brands, Florasis and Perfect Diary with a particular focus on their advertisements on overseas online platforms. The comparative study reveals that there are significant differences in the brand images established by the two brands in overseas markets. The advertisement of Florasis has distinctive Chinese cultural characteristics and expresses the concept that "the national is the global", which reflects the unique cultural specificity in the international market. In contrast, Perfect Diary's advertising is integrated into the international market without Chinese traces, which shows its ambition to compete with other world-renowned brands.

Objective: The goal of the current research is to identify the dominant phytochemical from the plantNyctanthesarbor-tristis Linn. and to investigate their binding affinities against the proteins BRaf Kinase mutant (3OG7) and Hsp90 Chaperone (2VCJ) that causesmelanoma. Methods: In this work, Schrodinger software was utilized to investigate the anti-cancer potential of phytochemicals Nyctanthesarbor-tristis against specific target proteins, namely BRaf Kinase mutant (3OG7) and Hsp90 Chaperone (2VCJ) Inhibitors. Results: Based on the outcome of the docking investigation, phytochemicals that exhibited highest binding affinity to the specified protein targets were subjected to induced fit docking and MM-GBSA computations using the Schrodinger Maestro version 2021.2 in prime module. According to the analysis, the compounds with the highest binding affinities for 2VCJ and 3OG7 are Arbortristoside D and Nicotiflorin respectively. The compound that interacted with both the proteins wasArbortristoside B. These phytochemicals appear to be more effective to the FDA-approved V600E-BRaf inhibitor Vemurafenib and Hsp90 Chaperone Inhibitor Diclonine. Conclusion: One of the most common, deadly, and dangerous malignant diseases with a high global prevalence rate is melanoma (skin cancer). The present study may prove more helpful in developing an ideal targeted drug delivery system of phytochemicals obtained from plant Nyctanthesarbor-tristisfor treatment of melanoma. This suggests that these substances could be evolved into highly effective anti-melanoma drugs.

In this study, we set out to measure race and gender bias prevalent in text-to-image (TTI) AI image generation, focusing on the popular model Stable Diffusion from Stability AI. Previous investigations into the biases of word embedding models—which serve as the basis for image generation models—have demonstrated that models tend to overstate the relationship between semantic values and gender, ethnicity, or race. These biases are not limited to straightforward stereotypes; more deeply rooted biases may manifest as microaggressions or imposed opinions on policies, such as paid paternity leave decisions. In this analysis, we use image captioning software OpenFlamingo and Stable Diffusion to identify and classify bias within text-to-image models. Utilizing data from the Bureau of Labor Statistics, we engineered 50 prompts for profession and 50 prompts for actions in the interest of coaxing out shallow to systemic biases in the model. Prompts included generating images for "CEO", "nurse", "secretary", "playing basketball", and "doing homework". After generating 20 images for each prompt, we document the model’s results. We find that biases do exist within the model across a variety of prompts. For example, 95% of the images generated for "playing basketball" were African American men. We then analyze our results through categorizing our prompts into a series of income and education levels corresponding to data from the Bureau of Labor Statistics. Ultimately, we find that racial and gender biases are present yet not drastic.

Abstract Science identity, or one's sense of recognition and competence as a scientist, is an invaluable tool for predicting student persistence and success, but is understudied among undergraduates completing preparatory work for later studies in medicine, nursing, and allied health (“pre‐health career students”). In the United States, pre‐health career students make up approximately half of all biology students and, as professionals, play important roles in caring for an aging, increasingly diverse population, managing the ongoing effects of a pandemic, and navigating socio‐political shifts in public attitudes toward science and evidence‐based medicine. Pre‐health career students are also often members of groups marginalized and minoritized in STEM education, and generally complete their degrees in community college settings, which are chronically under‐resourced and understudied. Understanding these students' science identities is thus a matter of social justice and increasingly important to public health in the United States. We examined science identity and engagement among community college biology students using two scales established and validated for use with STEM students attending four‐year institutions. Exploratory and confirmatory factor analysis were used on two sub‐samples drawn from the pool of 846 participants to confirm that the factor structures functioned as planned among the new population. Science identity values were then compared between pre‐health career students (pre‐nursing and pre‐allied health) and other groups. Pre‐health career students generally reported interest and performance/competence on par with their traditional STEM, pre‐med, and pre‐dentistry peers, challenging popular assumptions about these students' interests and abilities. However, they also reported significantly lower recognition than traditional STEM and pre‐med/dentistry students. The implications for public health, researchers, and faculty are discussed.