Not The Dreaded Doom

The first time I heard of and vaguely understood the phrase “machine learning” was roughly a year ago in a piece featured by Science Daily. I am loosely summarizing when I say that the point was that they were using images and computer algorithms to determine that it was time to harvest lettuce. This was mildly disappointing for me, as  when I read “machine learning”, I found myself thinking of SkyNet from the titular film, “The Terminator”. However, the thought of computers being able to tell when it was time to cut that juicy piece of lettuce down on their own was equally fascinating.

“Terminator again!” by Andy Veilski is licensed under CC BY-NC 4.0

In the present, I am wrapping up my first week with RET2020 and my lovely cohort, and these conversations on Machine Learning are just beginning.

Continue reading →

Our Summer Web-Based Institute for Technologies in CompSci and Healthcare (SWITCH) had a great start this week. It was such a wonderful opportunity to be able to meet, collaborate and share with educators who shared my passion for Computer Science Education. 

The goal this week was to complete an informal literature review and be able to write and compile python code. We met (virtually) every day from 9:30 am – 11:00 am to discuss research readings and then spent two to three hours each day in tasks that would help us reach our weekly goal. During our morning meetings we also had visiting expert lectures which were quite insightful. 

For example, did you know that every 30 seconds, one American will be diagnosed with Diabetes and another will suffer a coronary event? 

Presentation on ML & Diabetes Research

Gerard Cote, Director of Paths-UP presented in detail, the work that the PATHS-UP program aims to achieve and how our program ties into influencing the next generation of innovators. 

ML in Healthcare

Both Gerard and Souptik Barua spoke about how Machine Learning can impact the field of Healthcare, its present challenges and future opportunities. It was a gift to be able to learn about the Engineering Research Center and some of the research currently being conducted to positively impact the health of underserved populations in our country. 

In the afternoon, I dedicated myself to learning about Python, primarily how it differs from Java, which I was already familiar with. I quickly found out that Python is very intuitive and has a user-friendly syntax, which explains why it’s a common entry point for many programmers! 

I also read the articles for our informal literature review. It was truly impressive to learn about Computer Vision and am more fascinated than ever to learn more about it!

    I am proud to be a member of the SWITCH – PATHS-UP research program. For the first time, the SWITCH program will be completely virtual. Week one has been a whirlwind. I cannot believe we are one week down.

    Monday, we all met for the first time. Zoom is our go-to virtual environment. Our first assignment was a reading list. The reading list contained articles on Machine Learning and Computer Vision. I was amazed at how much we rely on Machine Learning and Computer Vision in our daily lives. Machine Learning and Computer Vision have grown exponentially in recent years. Both will play a significant role in our “Gaze Detection” project.

Detecting attention levels

    One of the essential things (among others) I gleam from programs such as this is the chance to learn from other educators. The program has a vast collection of educators with impressive backgrounds. I am sure I will learn a great deal from this team.  During our Zoom meeting, we had the opportunity to meet and hear Dr. Ashutosh Sabharwal. He discussed the fantastic research that is taking place in the field of Diabetes. 

      Wednesday, we had a discussion session that contained a Q & A with Gerry Cote from Texas A & M. He gave an overview of the PATHS-UP program. The Grand Challenges the program is trying to accomplish is assisting underserved populations with low SES with Diabetes education. The program aims to create a paradigm shift in the perception of healthcare in underserved communities.  

Areas of concentration for Diabetes Research

    They are working on a cost-effective device to monitor blood levels in individuals that are Diabetic. Dr. Cote mentioned  two devices that are in use today the “Lab on your waist” and the “Lab on your Palm.” These devices help people unobtrusively monitor glucose levels for those with Diabetes. Unfortunately, these devices are expensive to obtain and maintain for persons of low SES. New, more affordable devices are in development.  

    Thursday, we met with Mary Jin. She is a graduate student at Rice University. Mary is one of the leads in the program. She brings an enormous amount of energy, research experience as well as technical knowledge to our team. She introduced us to OpenCV. OpenCV is a program that allows programmers to capture images and video files from computer webcams; this is very exciting! Stay tuned! 

Week One

We have officially completed week one! I didn’t know what to expect coming into this program so I prepared by doing the Coursera Machine Learning course by Andrew Ng. That course is no joke at all! He jumps right into supervised/unsupervised learning, cost function, hypothesis. When the easiest thing you cover is linear algebra you know it’s serious work.

Before the official first day I spent the night reading the provided articles to try to get some background knowledge on Machine Learning. The first article I read was eye opening! A few of the things in there I had heard about but I had no idea how useful machine learning could be in agriculture. The concept of a drone being able to identify weeds and spray them with pesticide blows my mind and my wife was shocked to hear about Pinterest Lens. We had no idea that existed!

Our first presenter was Mr. Ashutosh Sabharwal and he told us about his research into Sociability. His research showed that there might be a correlation between how sociable one is (the amount daily steps taken, average text message length, and the amount of text messages) and depression.

Our second presenter Souptik showed us how machine learning could improve the people self-manage diabetes.  I didn’t realize how uninformed I was about living with diabetes. The fact that a person can get an inaccurate glucose reading depending on the time they took a sample is a little bit scary to me. Luckily with Souptiks research hopefully he can help change that!

Last but not least we finished off the week with a presentation from Mr.Gerard L. Cote from A&M who talked to us about the idea of having cheap, easy to wear, reliable tech that can take readings from patients without them having to drive to a hospital. It really is timely seeing as this quarantine period has made face to face appointments risky!

This week was extremely rewarding and I already moved 2 articles into my teaching folder so I can use them with my class this fall! Very excited to see what else we learn from this program! 👊