41530 research outputs found
Sort by
An Analysis Of Indoor Air Quality At Cal Poly For Sensor Design
Get PDFPrior research has shown that indoor air quality (IAQ) impacts cognitive performance. At Cal Poly, many older buildings are unable to maintain appropriate IAQ because of their outdated ventilation systems and the increasing number of students in the rooms. This work analyzes the IAQ of different buildings at Cal Poly, with a focus on Building 20. Carbon dioxide, temperature, and relative humidity inside classrooms are collected using an integrated circuit sensor and a microcontroller. A total of 38 hours of data was collected, with 22 of those hours in Building 20 specifically. We find that unlike temperature and relative humidity, CO2 levels routinely exceed 1,000 ppm—a concentration that hinders cognitive function. A questionnaire distributed to Cal Poly students suggests that while students can recognize poor IAQ in classrooms, they erroneously attribute these poor conditions to temperature and humidity instead of CO2. This data is then used to propose a system which can collect long-term data based on optimal placement, storage, and power requirements
Development Of An Advanced Step Counting Algorithm With Integrated Activity Detection For Free Living Environments
Get PDFPhysical activity plays a crucial role in maintaining overall health and reducing the risk of various chronic diseases. Step counting has emerged as a popular method for assessing physical activity levels, given its simplicity and ease of use. However, accurately measuring step counts in free-living environments presents significant challenges, with most activity trackers exhibiting a percent error above 20%. This study aims to address these challenges by creating a machine learning algorithm that leverages activity labels to improve step count accuracy in real-world conditions. Two approaches to balancing data were used: one employed a simpler oversampling technique, while the other adopted a more nuanced approach involving the removal of outliers. Models 1 and 2 were trained on each of these uniquely balanced datasets. Model 1 performed much better than Model 2 on testing datasets, but both achieved better than 20% error on new datasets, indicating their potential for more accurate step counting in real-world conditions. Despite challenges such as data imbalance, the study demonstrated the viability of using activity labels to enhance step counting accuracy. Future research should focus on addressing data imbalances and exploring more advanced machine learning techniques for more reliable activity monitoring
Feminist pedagogical requirements of vulnerability in writing failure: Bad claims and worse reviews
Get PDFOur students have limited perception of their faculty based on our classes, or for our graduate students in our publications. This is a potentially harmful for emerging scholars to understand faculty work only in its final form. Feminist scholars need to actively seek spaces in their courses to note the challenges and limitations of their own work[1]. In this critical commentary, I offer an overview of student standpoints, feminist possibilities and futures with less threating failures. The emotional and face threating vulnerability in failure by feminist scholars can create more welcoming failures moving forward
Failure to Contain or Failure of Imagination? Exploring Artificial Intelligence’s Implications for Feminist Pedagogy
Get PDFA User-Centric Continuous Authentication Modality Evaluation And Selection Scheme
Get PDFOne of one-time authentication’s most prominent vulnerabilities is the possibility of lunchtime attacks. In such scenarios, an adversary could exploit an unattended device with an active session and no measures are taken to prevent them from committing malicious acts. To address this issue, continuous authentication is utilized by continuously verifying whether an individual is a device’s rightful owner through various modalities, with data sourced from sensors. In the current body of research within this rising domain, various single-modal and multi-modal continuous authentication systems exist, that focus on employing unique combinations of modalities and improving existing supervised learning models used to solve this classification problem. However, no solutions allow prospective continuous authentication users to obtain the most suitable combination of modalities given their unique circumstances. Therefore, in this thesis, we design a user-centric continuous authentication modality evaluation and selection scheme. The scheme employs a multi-criteria decision analysis model, which involves compiling a list of continuous authentication systems, modalities, and associated sensors. As part of this scheme, we design security, privacy, and usability frameworks to conduct systematic analyses of the list of sensors and modalities gathered, while considering the system’s performance. The proposed scheme can be utilized to generate a ranked list of combinations of modalities appropriate for the user
Causal Inference Using Bayesian Network For Search And Rescue
Get PDFPeople who are considered missing have much higher probabilities of being found dead compared to those who are not considered missing in terms of Search and Rescue (SAR) missions. Dementia patients are incredibly likely to be declared missing, and in fact after removing those with dementia the probability of the mission being regarded as missing person case is only about 10%. Additionally, those who go missing are much more likely to be on private land than on protected areas such as forests and parks. These and similar associations can be represented and investigated using a Bayesian network that has been trained on Search and Rescue mission data. By finding associations between factors that affect these missions, SAR teams can find patterns in historical cases and apply them to future cases in order to narrow down their search areas, improve their plans, and hopefully lead to lower search times and fewer deaths and unsolved cases. Causal inference allows causal relationships to be determined, telling SAR teams that they can make current decisions based on these learned relationships and their decisions will cause the change that they expect based on the Bayesian network
Feedback Control System Analysis and Simulation for Laboratory Development
No full textControl System Theory has essential applications in Electrical Engineering, requiring a solid understanding of concepts like transfer functions, transient response, feedback, and stability. At California Polytechnic State University, San Luis Obispo (Cal Poly, SLO), the emphasis on learn-by-doing and hands-on experience is key to the educational approach. This analysis investigates the potential of updating Cal Poly’s Classical Control Systems Laboratories by analyzing and comparing virtual curriculum offerings, control education at Cal Poly, and curriculum at other universities. In addition to an analytical approach, some Quanser Inc.Ⓡ computer simulation examples are provided for assessment
Impact Force Helmet Device for Concussion Monitoring
Get PDFTo track head trauma in football, our device is placed inside a football helmet to measure both linear and gyroscopic (rotational) accelerations. These measurements are taken using an ADXL375 linear accelerometer and an LSM6DSOX accelerometer and gyroscope. The data from these sensors is processed and transmitted wirelessly by a SparkFun Thing Plus microcontroller to another SparkFun Thing Plus microcontroller connected to a laptop on the sidelines. Our program on the laptop stores the acceleration impact data into a .csv file for each player. If a likely concussive impact is detected, the program displays a popup warning. Testing results were promising and generally successful, although the ADXL375 accelerometer often produced inconsistent and inaccurate readings
High Frequency: A Graphic Anthology of Music
Get PDFHigh Frequency experiments with the production of both non-traditional book design and graphic design in order to explore the personal relationship that can exist between an individual and music. Key areas of experimentation lie in typography, graphic design, and layout in order to result in a specialized, hand-crafted book featuring custom posters of twenty-five albums that have had particular significance to my personal musical journey. How can one visually represent their emotional connection, journey, and relationship to a body of music? Though this is the primary focus of this project, a secondary goal is to explore how an individual interacts and moves throughout the book in order to push the boundaries of design and reflect the movement and dynamism of music. High Frequency is in essence a visual representation of my positioning as a creative between the worlds of music and design — particularly, how the former influences and shapes the outcomes of the latter
Generative Data Augmentation: Using DCGAN To Expand Training Datasets For Chest X-Ray Pneumonia Detection
Get PDFRecent advancements in computer vision have demonstrated remarkable success in image classification tasks, particularly when provided with an ample supply of accurately labeled images for training. These techniques have also exhibited significant potential in revolutionizing computer-aided medical diagnosis by enabling the segmentation and classification of medical images, leveraging Convolutional Neural Networks (CNNs) and similar models. However, the integration of such technologies into clinical practice faces notable challenges. Chief among these is the obstacle of acquiring high-quality medical imaging data for training purposes. Patient privacy concerns often hinder researchers from accessing large datasets, while less common medical conditions pose additional hurdles due to scarcity of relevant data. This study aims to address the issue of insufficient data availability in medical imaging analysis. We present experiments employing Deep Convolutional Generative Adversarial Networks (DCGANs) to augment training datasets of chest X-ray images, specifically targeting the identification of pneumonia-affected lungs using CNNs. Our findings demonstrate that DCGAN-based generative data augmentation consistently enhances classification performance, even when training sets are severely limited in size