Respected sir these two professors have literally no idea what they are talking. They just make opinion. 1. They have confirmation bias in there statement " we have done lots of de-bias in model, we work with more than 200 company" when you ask them question about de-bias model. 2. They reverse their own points, first they said algorithm are not bias and when you ask same question "you say algorithm are not bias" then they reverse there point. 3. Ask them what is difference between "identify", "diagnosis" and "detect". These are medical term and people use them frequently without understanding differences between them. 4. He just compare global north with global south without understanding internals. 5. In X-Ray report you have image and you can easily detect cracks, lumps etc even with naked eyes. It is true image is a collection of pixels but it is not about single pixels. Apart from that deep learning model is black box in nature and defining output is very difficult because there are multiple layer of neurons and each layers have multiple neurons. There are highly complex mathematical equation operate behind the neural network. Right now we have some algorithms and libraries which explain model results but still there is on going research in these field (Explanable AI). Both of these professors just advertise there own book.
Respected sir these two professors have literally no idea what they are talking. They just make opinion.
1. They have confirmation bias in there statement " we have done lots of de-bias in model, we work with more than 200 company" when you ask them question about de-bias model.
2. They reverse their own points, first they said algorithm are not bias and when you ask same question "you say algorithm are not bias" then they reverse there point.
3. Ask them what is difference between "identify", "diagnosis" and "detect". These are medical term and people use them frequently without understanding differences between them.
4. He just compare global north with global south without understanding internals.
5. In X-Ray report you have image and you can easily detect cracks, lumps etc even with naked eyes. It is true image is a collection of pixels but it is not about single pixels.
Apart from that deep learning model is black box in nature and defining output is very difficult because there are multiple layer of neurons and each layers have multiple neurons. There are highly complex mathematical equation operate behind the neural network. Right now we have some algorithms and libraries which explain model results but still there is on going research in these field (Explanable AI).
Both of these professors just advertise there own book.
Thank you for watching and sharing your views.