Mansi Singhal · Transcript · · Updated
Next 36 Application Video - Mansi Singhal – Transcript
Video by Mansi Singhal · Watch the original on YouTube
Summary
AI-generated from the transcript. Verify details against the source video.
- Mansi Singal studied 24 different Alzheimer's data sets over the past 2 years and found that every single data set was biased.
- Researchers and industry leaders around the world spend hours with this data, yet their models fail when tested on different ethnic groups.
- Mansi Singal is building a data integrity platform that provides unbiased data sets and saves time for universities, industrialists, and hospitals.
Use this video with Pro
Reading, copying and transcript downloads are free. Create saved AI results for this video with a signed-in Pro account ($4.99/month).
Transcript
Search ignores accents. All transcript text stays visible.
Hi, my name is Mansi Singal and I'm a PhD student at the University of Calgary. So, as we know AI mostly used in health care nowadays, but the question no one wants to ask is is the data we are using actually representative of every population we serve? Think about that because in Canada we are home to indigenous communities, Asians, Africans, and people from every corner of the world
and different populations respond differently to the same disease. The data should reflect that, right? But it does not. I studied myself for 24 different Alzheimer's data set over the past 2 years. Every single data was biased and I'm not alone. Researchers, industry leaders around the world has spent hours and hours with this data yet their models fail when tested on different ethnic groups. And this is dangerous
because a wrong prediction in health care is not just an error. It can harm or even kill the right patient. So, that is why I'm building a data integrity platform that provide unbiased data set and save time to university, industrialists, and hospitals so that they can build AI that work for every population and not just one. Thank you.