Difference between revisions of "User:ANoecker"
(→Executive Summary) |
|||
| Line 4: | Line 4: | ||
{| cellpadding="10" cellspacing="10" | {| cellpadding="10" cellspacing="10" | ||
| − | | [[Image: | + | | [[Image:Final-Poster-Submission-Andrew-Noecker.png|200px|link=http://www.mscs.mu.edu/~brylow/reu/2018/Posters/Poster-2018-Coffee-Johnson.pdf]] |
| '''[[Intrusion Detection in Swarm Robotics]]''': <br /> Student Researcher: [[Lindsey Coffee-Johnson]]. <br /> Mentor: Dr. Debbie Perouli. | | '''[[Intrusion Detection in Swarm Robotics]]''': <br /> Student Researcher: [[Lindsey Coffee-Johnson]]. <br /> Mentor: Dr. Debbie Perouli. | ||
[https://doi.org/10.1109/HRI.2019.8673263 Detecting Anomalous Behavior of Socially Assistive Robots in Geriatric Care Facilities]<br /> | [https://doi.org/10.1109/HRI.2019.8673263 Detecting Anomalous Behavior of Socially Assistive Robots in Geriatric Care Facilities]<br /> | ||
| Line 11: | Line 11: | ||
|- | |- | ||
|} | |} | ||
| − | |||
== Project Description== | == Project Description== | ||
Revision as of 19:14, 5 August 2021
Contents
Executive Summary
As a double major in Math and Computer Science with a concentration in Statistics and Data Science, participating in Marquette University's Research Experience for Undergraduate Students (REU) will develop necessary skills for the future. These include conveying and communicating the importance of the research in tandem with data management, visualization, and architecture. This will help fill the gap of data science researchers who can think creatively to solve problems in interdisciplinary fields such as healthcare and nutrition. A collaborative team including myself and mentorship from Dr. Bialkowski and Dr. Gretebeck will towards Capturing nutritional value at the point of consumption using accessible and inexpensive technologies.
|
Intrusion Detection in Swarm Robotics: Student Researcher: Lindsey Coffee-Johnson. Mentor: Dr. Debbie Perouli. Detecting Anomalous Behavior of Socially Assistive Robots in Geriatric Care Facilities |
Project Description
The National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) seeks to optimize metabolic response in individuals or population subgroups through tailored dietary approaches to promote health and prevent and treat disease. Accurate assessment of nutritional intake among community-dwelling individuals at the point of intake is a major obstacle in precision nutrition evaluation. Without this fundamental information about nutritional lifestyle factors, a highly modifiable environmental condition, disparities across Race, Ethnicity, socioeconomic status, and age will persist. Our team of cross-disciplinary researchers is combining expertise in hyperspectral imaging, image processing, mobile technology, analytics, nutrition, psychology and human motivation to develop a robust and accurate platform of nutrition evaluation at the point of consumption. This important translational step forward will empower researchers across disciplines with the information needed to facilitate meaningful change and more equitable access to health.
Main Objectives
- Understanding and capturing the information that we're working on and being able to tell people why this is important and why it matters
- Develop robust skills in data management and data architecture
- Utilize visual analytics to communicate a story about this effort to capture nutritional value at the point of consumption
Weekly Log
Week 1: 6/1 - 6/4
- Attended orientation
- Read grant proposal in order to get up to speed on the project
- Met with mentors to discuss project
- Learned about and set up wiki page
- Developed list of desired skills that will be grown throughout the summer; these will be used to map out goals/milestones for the next 8-9 weeks in collaboration with Dr. Bialkowski and Dr. Gretebeck
- Performed literature review of available technology for:
- Capturing food images
- Quantifying food volume
- Prepared brief summary paper and concise presentation for mentors and overall project team
Week 2: 6/7 - 6/11
- Completed Responsible Conduction of Research Training for Federal Researchers (synchronous with Dr. Brylow on Monday, asynchronous modules on Tuesday)
- Performed Literature Review on technology currently available for quantifying food nutrient composition, specifically examining hyperspectral imaging
- Met with mentors to discuss how hyperspectral imaging is used for crops and harvest
- Discussed opportunities to use this technology at the point of consumption
- Met with REU fellow from 2020 to hear about her experience and advice she had for the program
Week 3: 6/14 - 6/18
- Reviewed and characterized codebook of USDA Food and Nutrient Database for Dietary Studies (FNDDS)
- Created draft of new codebook for convolutional neural network database storing information about:
- Image information
- Segmented food volume
- Nutrient content
- Designed paper prototype for poster presentation
Week 4: 6/21 - 6/25
- Read about data gathered in other hyperspectral imaging studies
- Continued to work on drafting codebook by:
- Updating format and style
- Adding variables that merge with BAP
- Adding hyperspectral image sensor variables
- Met with Dr. Daniel Pinto and his graduate students to discuss the following regarding Brief Action Planning (BAP):
- What it entails
- How data is stored and what variables are examined
- How it can be transferred to nutrition from physical activity
- Began to clean USDA Food and Nutrient Database for Dietary Studies (FNDDS) and create some preliminary visuals
- Researched target wavelengths for proteins, fats, and carbohydrates
Week 5: 6/28 - 7/2
- Prepared and gave lightning talk on progress thus far for REU program
- Researched appropriate wavelengths for targeting fats, proteins, and carbohydrates
- Continued cleaning USDA FNDDS and brainstorming possible visuals
- Prepared several visuals using FNDDS that could inform users when making decisions about food
Week 6: 7/5 - 7/9
- Began reading on how to simulate data in R
- Modified prepared visuals following suggestions from project team at 7/2 meeting
- Continued creating visuals using FNDDS to tell a story about the data
- Isolated breakfast foods and learned about waterfall and balloon plots as a way to quickly compare different breakfasts
- Isolated fruits and vegetables and created a heat map to quickly analyze different nutrient values
Week 7: 7/12 - 7/16
- Created new waterfall plots and refined ones from last week
- Created new faceted bar plots showing different nutrient values for all breakfast foods
- Continued reading on simulation and began simulating nutrient values for food groups
- Simulated user confidence levels as a way to show how those could change over time
- Attended Northwestern Mutual Industry Panel
Week 8: 7/19 - 7/23
- Simulated what an ideal NIR spectrum chart would look like for protein, fats, and carbs
- Updated faceted bar plots by reordering nutrients to tell a more comprehensive story
- Created plot that showcases monounsaturated, polyunsaturated, and saturated fats as a proportion of total fats
- Grouped foods into groups based of meal, then created faceted waterfall plot to show different food's fat content per half pound. This could be quickly replicated for any other nutrient
- Began drafting final poster
Week 9: 7/26 - 7/30
- Edited colors and shading on simulated NIR spectra plot and faceted waterfall plot to enhance the story telling
- Drafted a 48x36 poster to convey the work done this summer in an informative and visually appealing manner
- Presented the poster to the Nutrition Project Team and received critical feedback to help improve presentation skills and the poster itself
- Began outlining final research paper to be written next week
Week 9/10:
- Edited and submitted final poster
- Create final poster for presentation
