PRESENTERS

Department of Biological, Human & Life Sciences

Examining Correlations Between Creative Technological Design and Consumer Cognitive Dissonance

Our study explores cognitive dissonance using an innovative method, examining the contrast between consumer beliefs and behaviors in reaction to different creative technological product designs. Cognitive Dissonance is the discrepancy between one's external actions (behaviors) and their internal values (attitude). The research focuses on consumer behavior, in response to technological creative designs of AI, video games, autonomous vehicles, smart home technologies, and health care. Our study developed a questionnaire to assess consumer product market behavior associated with these various innovative technologies to measure conflicting views of the applications with these various technologies. Our preliminary analysis demonstrates potential correlation with adult demographics with autonomous vehicles and video game technological designs. The preliminary analysis also indicates potential gender-specific trends of dissonance upon creative AI based technologies in health care. 

RESEARCHERS: Eliana H., Monte Vista High School '25; Moksha R., Mission San Jose High School '25; Avigna S., Dublin High School '25; Arya S., Dougherty Valley High School '25; Maddy Z., Amador Valley High School '24; Alice Z., Amador Valley High School '25

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: Cognitive Dissonance | Technology | Consumer Behavior | Artificial Intelligence | Creative Design

Department of Computer Science & Engineering

Exploring Potential Solutions To The Small-Scale Problems of the Lambda Cold Dark Matter Model (λCDM)

The Lambda Cold Dark Matter (Lambda CDM), which depicts dark matter particles as cold and collision-less, is successful for large scale structures but has faced challenges on the small scale due to contradictions with astronomical observations. The challenges faced are called the small-scale problems, which are: the core-cusp problem, the missing satellites problem, and the too-big-to-fail problem. Alternate models, such as the Self Interacting Dark Matter (SIDM) model or the implementation of baryonic feedback effects, have been proposed to combat these small scale problems. Unlike Lambda CDM particles that primarily interact through gravity, SIDM particles do interact with each other through means of self-scattering processes. In this paper, we utilize recent simulations to explore the possibility of the SIDM model with baryonic feedback combating the small-scale problems while accurately representing the universe on the large scale to ultimately replace the Lambda CDM model.

RESEARCHERS: Sakash G., BASIS Independent Fremont '24; Samuel L., BASIS Independent Silicon Valley '25; Michael C., Los Gatos High School '25; Anvith K., Foothill High School '25; Prahlad Vangeepuram C., Mission San Jose High School '26; Ashwin M., Mission San Jose High School '27; Katrina R., Monta Vista High School '26

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Dark Matter | Cold Dark Matter | Self Interacting Dark Matter | baryonic feedback | small-scale problems | simulations

Department of Computer Science & Engineering

Using AI Pathfinding to Aid EMS

Using autonomous drones to autonomously scan an area after natural disasters (or other events) can be used to effectively make an algorithm that creates more efficient or alternative routes to aid EMS.This research project aims to study whether autonomous drones can effectively be used to scan and efficiently provide paths in natural environments. This project additionally aims to reduce the overall need for drone pilots as and decrease search time. While this project focuses on autonomous drones for search and rescue, the findings could beneficial for other uses such as delivery or surveillance.

RESEARCHERS: Aryav D., Mission San Jose High School '26; Tarun M., East Sacramento High School '25; Kaustubh L., Mission San Jose High School '26

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: AI | Drones | Ml | Autonomous Vehicles | 3d Printing

Department of Biological, Human & Life Sciences

A Comparative Analysis of Human and Language Model Creativity

Challenging the prevailing perception that artificial intelligence is far less creative than humans, our study explores the creativity of Large Language Models (LLMs) by focusing on their ability to impersonate individuals. Utilizing established measures such as the Alternate Uses Task (AUT), Remote Associates Task (RAT), and Torrance Tests of Creative Thinking (TTCT), we compiled an original creativity assessment that quantifies both convergent and divergent thinking. Furthermore, our test includes the self-reported personality measure of the NEO Five-Factor Inventory (NEO-FFI) to generate an extensive profile of the participant’s creativity. Moreover, in facilitating an unbiased comparison between the LLM and human participants, we provided the LLM with the participants’ demographic data, including but not limited to race, age, gender, profession, salary, and location. Such allowed for more personalized and human-like responses from the LLM, producing a reliable comparison between the two. Our preliminary data indicates that LLMs demonstrate significant creativity, outscoring humans on all three creativity assessments while offering novel, comprehensive, and relevant responses to each test. Our preliminary observations challenge previous assumptions regarding the creativity of AI and clarify its positive relation to humans. Our findings allow for further exploration into the dynamics between humans and AI as a means to bolster human creativity. 

RESEARCHERS: Samuel L., Valley Christian High School '25; Nishka V., Foothill High School '25; Lucy W., BASIS Independent Fremont Upper School '25

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: Creativity | Neuroscience | Large Language Models | Artificial Intelligence | HIPAA

Department of Biological, Human & Life Sciences

Determining Neural Correlates for Different Affective Responses to Musical Stimuli and Classifying Musical Affection through Machine Learning

More info to come

RESEARCHERS: Aryan K., Ayaan K, Maya S., Nishita V.

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: 

Department of Computer Science & Engineering

Which machine learning algorithm is s better for autonomous driving between Supervised Learning and Deep Reinforcement Learning

Autonomous driving has been a growing field in the car and technology industry and when used properly can be the safer and also more efficient choice to dependent driving. Autonomous driving can promote greater road safety, reduced road congestion, environmental gains and more productivity. Our objective is to identify the best machine learning method (supervised learning v.s. deep reinforcement) to train autonomous vehicles and thus achieve greater car safety. To compare the machine learning models’ performance, we built a miniature car (using raspberry pi and various sensors) modeled after an autonomous vehicle as well as a maze and obstacles. The machine learning method that leads to the car completing the maze the fastest, will be identified as the more efficient and accurate one. To train the Supervised Learning model, we are having the car go through the maze numerous times, and creating our own dataset based on values gathered. For the Deep Reinforcement model, we will have the car navigate the maze and make the decisions so the agent is able to learn and improve. 

RESEARCHERS: Mihir R., Archbishop Mitty High School '26; Devyn P., Saint Francis High School '25; Truman Y., Basis Independent Silicon Valley '25; Ishaan G., Fremont High School '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Autonomous Vehicles | Supervised Learning | Deep Reinforcement Learning 

Department of Biological, Human & Life Sciences

The effects of AMF on nitrogen-rich soil and normal soil in normal environment and an environment with the abiotic stress of high salinity of the soil.

Nitrogen is an essential component of chlorophyll, which is important in the process of photosynthesis. In addition, AMFs have a symbiotic relationship with plants and improve nutrient absorption and resistance to abiotic stressors. This fungal mycelium colonizes the roots of plants and specifically improves nutrient absorption by regulating the transfer of phosphorus and nitrogen in the soil. Plants also display a variety of effects when they are in soil with different salinity levels. In order for plants to survive and grow in soil with a medium and high salinity, it is crucial that they maintain proper osmotic balance. Without this, the cells within the plants may dehydrate and eventually die. Altered soil salinity can lead to issues with ion toxicity, osmotic stress, nutrient deficiency, and oxidative stress, which limits the uptake of water from the soil. Soil salinity also reduces the phosphorus levels in the plant, because Ca ions precipitate with phosphate ions. Many salts are also essential nutrients in the soil, so high levels of salt in the soil, or high salinity, can interfere with nutrient uptake. To remain the ideal balance and analyze the different growth of plants with the abiotic stress of high salinity on them, we will conduct many experiments and understand the impact AMF and nitrogen will hold because of it's resistance to such abiotic stresses in our environment. 

RESEARCHERS: Sai Deeksha M., Irvington High School '26, Gia G., Mission San Jose High School '26; Laasya C., Basis Independent Fremont Upper School '24; Neha A., Mountain House High School '25

ADVISOR: Kaur, Environmental Biology & Genetics

KEYWORDS: Sugar snap peas | Arbuscular Mycorrhizal Fungi | Nitrogen | Abiotic Stress | High Salinity | plant biology

Department of Computer Science & Engineering

Python-based Hartree-Fock approximations for molecular energy

The ground state energy of a molecule is very useful in verifying the feasibility and stability of a given molecule. This energy is difficult to compute exactly, but approximation methods can be used to obtain very close estimates of the energy. One such method, called Hartree-Fock, is simulated with a Python program called PySCF to calculate these energies. Our research aims to verify the accuracy and efficiency of this program and make it more accessible for use in the future.

RESEARCHERS: Allen L., Monta Vista High School '25; Jayden L., Monta Vista High School '25; Ram S., Liberal Arts and Science Academy '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Quantum Chemistry | Molecular Energy | Hartree-Fock | PySCF

Department of Computer Science & Engineering

Theoretical study of quantum dots in noise-resistant quantum key distribution systems

Quantum key distribution (QKD) systems hold the potential for secure communication by leveraging the laws of quantum mechanics. However, challenges arising from noise and environmental interactions can compromise the security and efficiency of QKD systems. This research aims to explore the theoretical aspects of utilizing quantum dots within noise-resistant QKD systems. By leveraging advanced simulation techniques on a high-performance computing cluster, we are investigating the impact of quantum dots on enhancing the security and reliability of QKD systems.

RESEARCHERS: Prahlad S., Washington High School '25; Leah U., Lynbrook High School '24; Nidhi P., Lynbrook High School '24; Steven B., Foothill High School '25; Xiangtuo C., Basis Independent Silicon Valley '26; Eashan S., Round Rock High School '27; Achintya P., Summit Tahoma High School '24

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Quantum Dots | Quantum Key Distribution (QKD) | Communication

Department of Chemistry, Biochemistry & Physics

Design, Synthesis, and Biological Evaluation of C-4 Ester Analogs of Podophyllotoxin

More Info to come

RESEARCHERS: Breanna L., Harriet C., Grace Y., Kimberly K.

ADVISOR: Njoo, Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: 

Department of Computer Science & Engineering

Advancing Environmental Mapping and Forest Health Assessments: Integrating Automated Asessments using Autonomous Drones

With the increasing concern for environmental conservation, there is a growing need for efficient methods of environmental mapping and forest health assessments. However, traditional methods employed by the U.S Forest Health Monitoring have faced controversy due to limited spatial resolution and integration of modern technologies. This research paper explores the application of machine learning algorithms in autonomous drones to conduct forest health assessments. Autonomous drones have the ability to collect timely, up-to-date data, which offers enhanced accuracy. This study focuses on using NIR cameras to capture Red-Green-Blue to evaluate appropriate metrics, such as the Normalized Difference Vegetation Index (NDVI), Leaf Area Index (LAI), and foliage color, to create tree classes and identify forest health indicators. This study provides a more efficient, accurate, and up-to-date approach to monitor and evaluate the well-being of forests—supporting ongoing efforts towards environmental management and conservation.

RESEARCHERS: Andrew D., Leland High School '26; Navya R., Dougherty Valley High School '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Autonomous Drones | Machine Learning | Forest Health Assessment | Environmental Mapping | Deep Learning

Department of Chemistry, Biochemistry & Physics

Optimizing the Synthesis of Novel Carmofur Analogs for In Vitro Evaluation as Dual-Purpose Inhibitors Of Human Acid Ceramidase and the SARS-CoV-2 Main Protease

More Info to come

RESEARCHERS: Amber L., Leland High School '24; Lexi X., Leland High School '24

ADVISOR: Njoo, Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: 

Department of Biological, Human & Life Sciences

Dietary patterns during COVID-19: Analyzing changes in consumption habits in adults before, during, and after COVID-19 lockdown

The COVID-19 pandemic and the resulting quarantine have caused drastic changes in an individual’s lifestyle. Dietary lifestyle, relating to an individual’s relationship with both food and exercise, is a major part of this change. We hypothesize that the pandemic caused participants to change their consumption patterns drastically, and that these changes have remained in place even after the lock down period ended. Our questionnaire, which borrows ideas from four standardized eating assessments (SCOFF, EAT-26, CET, and EDE-Q) will aim to test this in adult participants. The results of this study will provide a better understanding of one of the several changes that pandemic has brought upon us. 

RESEARCHERS: Avi T., Saratoga High School ' 25; Anushka K., Alison Montessori High School '25

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: COVID-19 Pandemic | Quarantine | Dietary Lifestyle | Consumption | Standardized Eating Assessments

Department of Computer Science & Engineering

Hybrid Quantum-Classical Graph Generative Adversarial Network for Generating Chemically Stable Molecules w/ Python Molecular Bench Marking Processes

Present-day drug discovery methods cost billions of dollars and take five to ten years on average. To mitigate the high costs and reduce time needed, researchers have begun utilizing various computational approaches to search for molecules from the chemical space, which can be on the order of 10^60 molecules. One promising approach involves deep generative models, Artificial Intelligence (AI) models that learn from nonlinear data by modeling the probability distribution of chemical structures and creating similar data points from the trends it identifies. These generative models can extract salient features that characterize the molecules, but often require a lot of memory and can be inefficient. Aiming for an even faster runtime and greater robustness when analyzing high-dimensional data, our project builds upon our previous proof-of-concept Hybrid Quantum-Classical Generative Adversarial Network (QGAN) and remodeled it to specifically generate molecular graphs (QNetGAN). By adapting the internal architecture to operate on molecular graphs, we hypothesized that the new structure would enforce bonding between atoms and create connected molecules, instead of the scattered atoms from before. After synthesizing several sample molecules, our model achieved an appreciable 47% success rate compared to the previous 2.3%. To further improve the accuracy of our generated molecules, we are implementing molecular geometry optimization algorithms and benchmarking processes, while iterating through our QNetGAN models to increase the stability and chemical feasibility of our molecules and the accuracy of our model.These results ultimately demonstrate the future potential of our QNetGAN and a promising path towards more efficient, cost effective drug development processes.

RESEARCHERS: Adelina C., Archbishop Mitty High School '24; Max C., Sir Winston Churchill Secondary School '25 (Victoria, BC Canada); Hasset M., Piedmont Hills High School '25; Linda C., Homestead High School '24; Diya J., The Quarry Lane School '25; Aditya P., The Quarry Lane School '27

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Drug Discovery | Quantum Computing | Machine Learning | Generative Adversarial Network

Department of Chemistry, Biochemistry & Physics

Fragment-based Discovery and Anti-Cancer Activity of a Novel 5-methylisoxazole Piperazylacrylate Covalent Inhibitor for Broad Targeting of Key Oncogenic Markers

Covalent inhibitors have reemerged in the field of medicinal chemistry as previous concerns of off-target side effects were addressed through analysis of the specific protein binding pockets to determine the inhibitor’s structure activity relationship (SAR). Acrylamide warheads have been identified to efficiently bind to cysteine residues and are prevalent in many FDA-approved covalent inhibitors such as Sotorasib against G12C mutant KRAS and Afatinib against Epidermal Growth Factor Receptor (EGFR). This study aims to analyze bioactive fragments in combination with acrylamide warheads and their SAR in improving covalent inhibition of oncoproteins. Heterocycles make up more than 85% of bioactive molecules, and the 5-methyl-isoxazole heterocycle fragment had been previously utilized in medicinal drugs such as the FDA approved Dicloxacillin used to treat bacterial infections. The shown bioactivity of this molecular fragment gives it potential to aid in covalent inhibition. Inspired by this, we developed a library of 12 isoxazole-based acrylamide inhibitors. Although the results of computer modeling suggested that our molecules would inhibit G12C mutant KRAS, these compounds were also found to be potent in in vitro antiproliferative activity in human colorectal cancer cells lacking the KRAS G12C mutation, suggesting that these molecules might act through an alternative pathway. Further inspection of computer models and docking simulations suggested that these compounds may covalently inhibit other cancer targets, such as EGFR. With these preliminary biological results in hand, we further investigate the primary pathway of action of these compounds through flow cytometry, transcriptomic analysis of key markers by RT-qPCR, and protein expression mapping.

RESEARCHERS: Polina B., Los Altos High School '24

ADVISOR: Njoo, Synthesis | Physical Organic Chemistry | Catalysis | Chemical Biology | Spectroscopy | Medicinal Chemistry 

KEYWORDS: Covalent Inhibition | Organic Synthesis | Structure-activity Relationship | Medicinal Chemistry 

Department of Biological, Human & Life Sciences

Physiological and Neurological Characterization of Long-Term COVID-19

Long-term COVID-19 has affected millions of people across America. Patients that tested negative months ago still face physiological and neurological effects with unknown timeframes. By surveying people who tested positive for COVID-19, we are able to connect the data to demographic information to discover the longevity of COVID-19 effects. Participant responses have been collected through a questionnaire on Jotform, a HIPPA complaint website. We utilized automated scoring and will use R to analyze the data collected. Every participant is between 18-55 years, fluent in English, is a U.S citizen or permanent resident, and has not suffered from a serious medical condition prior or during the COVID-19 infection time. We made categories to assess the severity and timeline of physiological, cognitive, and working memory symptoms, and scored them on a scale of 0-3 for severities and 0-4 for timeline. We put up flyers in libraries, sent emails to school faculty, asked parents, and posted it on Instagram for outreach. Important results were noted through our first thirty participants. Factors like depression were lower in every participant compared to working memory. Physiological symptoms such as cough and fatigue affected more people for a longer duration compared to the cognitive symptoms included in our questionnaire. By observing demographic data, the majority (~97%) of our participants tested positive for COVID-19 after receiving the vaccination, which has been taken into consideration when examining the results. We aim to expand our scope of participants to include a larger variety of racial backgrounds and vaccination statuses. 

RESEARCHERS: Rohan M. Mountain View High School '25; Yukta C., California High School '24; Rishika M., Presentation High School '24; Ameya R., Washington High School '26

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: COVID-19 | Fatigue | Long-term Symptoms | Memory Impairment | Cognitive Effects | Physiological Effects

Department of Computer Science & Engineering

Diagnosing different types of skin cancer using Machine Learning (CNNs)

More information to be posted soon!

RESEARCHERS: Karen G., Lynbrook High School '26; Aryan G., West High School '25; Aarush N., Mountain House High School '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: 

Department of Biological, Human & Life Sciences

Decoding Inner Speech with Brain-Computer Interfaces: A Study in EEG Analysis and Machine Learning

Brain-Computer Interfaces (BCI) detect brain signals and translate them into commands which are carried out by other devices. For people affected by neuromuscular disorders, BCIs can greatly improve their quality of life by restoring lost function. These neuromuscular disorders often impede an individual’s ability to communicate, thus presenting a need for BCIs that can interpret inner speech. Electroencephalography (EEG) is a standard noninvasive neuroimaging technique measuring electrophysiological responses in the brain produced by synced neurons. Recent improvements in machine learning have led to advances in detecting brain patterns present in EEG data, allowing more promising and reliable BCIs. In this project, we utilize a dataset consisting of EEG data of inner speech commands from 10 subjects. Through analysis of the data using and the application of machine learning, we aim to develop a model that can accurately interpret inner speech. 

RESEARCHERS: Shravani V., High Technology High School '24; Sarah L., Amador Valley High School '24; Krrish G, Amador Valley High School '24; Sushmita M., North Creek High School '24

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: Brain Computer Interfaces | Electroencephalography | Data | Machine Learning | Inner Speech | Neural Networks | Channel | Brain Signals

Department of Computer Science & Engineering

Using Quantum Neural Networks (QNNs) and the Mathematical Morphological Reconstruction Algorithm for brain tumor detection

Brain tumors affect millions around the world, so detection is critical to helping doctors determine treatment. Currently, radiologists manually identify tumors through MRI (Magnetic Resonance Imaging) scans; however, this poses several limitations: it creates a heavy reliance on the experience of radiologists, has become increasingly costly and time-consuming, and is not as accessible to areas that lack the necessary resources and doctors. With the advancement of deep learning algorithms, a more accessible and efficient solution is possible. Given the existing research in classical Convolutional Neural Networks (CNNs) for tumor detection, Quantum Convolutional Neural Networks (QCNNs) offer a promising approach to the problem. Mathematical Morphological Reconstruction (MMR), another machine learning image processing method, provides a relative metric for success in the QCNN, and is another classical alternative to CNNs. This research compares the accuracy and computational speed of the MMR, QCNN, and CNN algorithms to determine whether introducing a quantum aspect presents any noticeable advantage. To build these models, extensive datasets of MRI brain scans were collected. The MMR algorithm involved applying various techniques such as dilation, erosion, and skull stripping through OpenCV2’s morphology functions. The QCNN algorithm utilizes the power of quantum to encode the data into a parametrized quantum circuit and apply convolutional and pooling layers. So far, our results indicate that the MMR algorithm achieved up to 92% accuracy. These results will be compared with the accuracy of the QCNN and CNN algorithms.

RESEARCHERS: Tiffany L., The Quarry Lane School '25; Riddhi S., Evergreen Valley High School '26; Eesha G., The Quarry Lane School '26; Shivansh B., Dublin High School '26; Rishav S., Dougherty Valley High School '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Mathematical Morphological Reconstruction (MMR) | Quantum Neural Networks (QNNs) | Convolutional Neural Networks (CNN) | Brain Tumor Detection | Quantum Computing 

Department of Biological, Human & Life Sciences

CovidVacMap: A Global Network Analysis of COVID-19 Vaccine Distribution to Predict Breakthrough Cases Among the Vaccinated Population

The COVID-19 pandemic, caused by the SARS-CoV-2 virus, has had a significant impact on the lives of many people, with numerous deaths and hospitalizations. In response, several vaccine manufacturers, such as Pfizer, Moderna, and Johnson & Johnson (J&J), have developed and tested COVID-19 vaccines. However, it is unlikely that society will achieve full immunity to COVID-19 due to the emergence of variants such as Omicron and Delta that may not be fully covered by current vaccines. The presence of multiple variants also means that current vaccines may not be as effective in preventing outbreaks. To help countries prepare for future outbreaks, the CovidVacMap project uses modeling to predict the risk of COVID-19 spreading across the globe.

RESEARCHERS: Pragyaa B., Basis Independent Silicon Valley '25;  Eddie Z., Harker School '26; Anaisha D., Dougherty Valley High School '24

ADVISOR: Jahanikia, Neuroscience

KEYWORDS: COVID-19 Pandemic | Network Science | Machine Learning | Data Science | Outbreaks

Department of Chemistry, Biochemistry & Physics

Evaluating Pharmacokinetic (ADME) Properties of Small Molecules

Understanding pharmacokinetics (ADME) is important to evaluate the safety and efficacy of drug molecules in biological systems. Our lab focuses on studying the absorption, distribution, metabolism and excretion (ADME) aspects of selected molecules by running in vitro experiments in the lab. Currently, selected molecules such as naringin, quercetin, and apitolisib are being investigated in our lab to evaluate its lipophilic behavior to understand its absorption and distribution aspects, its permeability aspects are being investigated to understand its transport across various physiological barriers, its plasma protein binding estimates the availability of drugs at the target site. The metabolic stability of the molecules will be evaluated to understand its excretion pathways. The results obtained from these assays help in optimization of drug molecules for its safe and effective delivery to the target site. 

RESEARCHERS: Reva U., Cupertino High School '26; Abir B., Dougherty Valley High School '26; Tulika S., Amador Valley High School '26

ADVISOR: Solgatra, Analytical Biochemistry & Phamarcokinetics Lab

KEYWORDS: ADME | Poly-Glycoprotein | Bioenhancer | Dual-inhibitor | Small Molecules 

Department of Biological, Human & Life Sciences

Quantifying the effectiveness of tide pool MPAs by comparing biodiversity indexes of MPAs with non-MPAs in Half-Moon Bay

Our group sought to investigate the impact of human activity on the Northern California rocky intertidal ecosystem. Our research centered around the question: How has human interaction with rocky intertidal zones affected their ecosystem health, as well as species biodiversity and abundance, when compared between protected and non-protected marine areas? Due to humans being able to take species from non-protected sites including but not limited to: moon snails, shore crabs, rock crabs, limpets, turban snails, sea urchins, mussels, oysters, hermit crabs, and even octopuses, the ecosystems of the non-protected tidepools have likely been severely affected. For example, because of predatory species being removed by humans, a possible effect is that many species of seagrass and other organisms in turn grew dramatically due to the uncompetitive environment. In addition, overfishing is largely unregulated and reduces the population count of organisms in the intertidal community. To support our hypothesis and the reality of the issue, we first went to Maverick’s Beach, a non-protected area. There, we took quadrat data using random sampling and transect lines. We collected data from Maverick’s Beach and Fitzgerald Marine reserve marine protected area through multiple trips.

RESEARCHERS: Vicki C., Mission San Jose High School '24; Tejin M., Athletin High School '26; Melody E., Tesoro High School '24

ADVISOR: Adams, Ocean and Marine Science

KEYWORDS: Marine Protected Area | Biodiversity | Tide Pools | Simpson’s Diversity Index | Intermediate Disturbance Hypothesis

Department of Computer Science & Engineering

Explainable AI Model to Detect Pneumonia from X-ray Images

Explainable Artificial Intelligence (XAI) refers to the practice of designing and developing AI models and systems in a way that their predictions, decisions, and behaviors are understandable and transparent to humans. Our project aims to leverage XAI techniques to enhance the interpretability of Convolutional Neural Networks (CNNs) for the analysis of lung images, facilitating a more transparent and insightful examination of respiratory conditions. In this project we used explainable algorithms such as LIME Saliency Maps, Grad-CAM, Occlusion Tests, Deep Taylor Decomposition and SHAP to visualize CNN decision areas and assess the features important for decision making. We successfully adapted established eXplainable Artificial Intelligence (XAI) methods to our models. Applying these XAI approaches to X-Ray Images enabled mostly successful differentiation between pneumonia-affected lung images and normal, healthy lung images. Our models offered a unique perspective by visually highlighting the specific areas within the images that influenced the Convolutional Neural Networks' decision-making process, shedding light on their inner workings.

RESEARCHERS: Vincent F., Basis Independent Silicon Valley '25;  Rohin V., Mission San Jose High School '26; Saahitya V., Mission San Jose High School '25; Anish B., Foothill High School '25

ADVISOR: Liu, Chemistry & Computer Neurobiology & Explainable AI & Augemented Reality

KEYWORDS: Explainable AI | Pneumonia

Department of Computer Science & Engineering

Theories to Describe Dark Matter

The Lambda Cold Dark Matter (λCDM) Model, which depicts dark matter particles as cold and collisionless, is successful for large scale structures but has faced challenges on the small scale due to contradictions with astronomical observations. The challenges faced are called the small-scale problems, which are: the core-cusp problem, the missing satellites problem, and the too-big-to-fail problem. Alternate models, such as the Self Interacting Dark Matter (SIDM) model or the implementation of baryonic feedback effects, have been proposed to combat these small scale problems. Unlike λCDM particles that primarily interact through gravity, SIDM particles do interact with each other through means of self-scattering processes. In this paper, we utilize recent simulations to explore the possibility of the SIDM model with baryonic feedback combatting the small-scale problems while accurately representing the universe on the large scale to ultimately replace the λCDM model.

RESEARCHERS: Sakash G., BASIS Independent Fremont '24; Samuel L., BASIS Independent Silicon Valley '25; Anvith K., Foothill High School '26; Michael C., Los Gatos High School '25

ADVISOR: McMahan, Quantum Computing & Computer Science 

KEYWORDS: Dark Matter | Cold Dark Matter | Self Interacting Dark Matter | baryonic feedback | small-scale problems 

Department of Biological, Human & Life Sciences

Shoot Proliferation of Some Fruit Ex-plants

Totipotency is the ability of all living plant cells to regenerate into an entire plant, and this serves as the foundational concept for plant tissue cultures and micropropagation. The primary focus of tissue cultures is to rapidly proliferate fruit plants, a process that will be vital for genetic transformation in the future. Developing meticulous explant sterilization protocols and creating tailored phytohormone concentrations in the media for different plant varieties is crucial to ensure successful tissue cultures.

RESEARCHERS: Pranati M., Quarry Lane School '25; Shambhavi S.,Amador Valley HS '24; Ram Rishi P., California HS '24; Amanda L., Piedmont Hills HS '24

ADVISOR: Poudyal, Biotechnology & Agronomy

KEYWORDS: Plant Biology | Tissue Culture | Proliferation