Hi,
Basically, my school is doing a science fair for the next two weeks (where we'll have to present to PhD students and professors at the university!), and the topic is relations between humans and AI. I was looking around and thought that the Breast Cancer Diagnosis project for middle school would be such a cool idea, but when I tried to follow the steps myself in the tutorial and Google Colaboratory, I found that the code wasn't working and there were some errors in the code. Could someone please help? I'd be super grateful.
Using the Breast Cancer Diagnosis Project for a Science Fair
Moderators: AmyCowen, kgudger, bfinio, MadelineB, Moderators
-
bfinio
- Expert
- Posts: 761
- Joined: Mon Aug 12, 2013 2:41 pm
- Occupation: Science Buddies Staff
- Project Question: Expert
- Project Due Date: n/a
- Project Status: Not applicable
Re: Using the Breast Cancer Diagnosis Project for a Science Fair
Hi,
Could you please post the exact steps you are taking and the error messages you receive when running the code? If you do that, we should be able to help.
Thanks,
Ben
Could you please post the exact steps you are taking and the error messages you receive when running the code? If you do that, we should be able to help.
Thanks,
Ben
Re: Using the Breast Cancer Diagnosis Project for a Science Fair
Thank you for the response! I realized I forgot to play one of the code blocks, which is why the rest didn't work. Although I do have another question: In the last two code blocks, when you change the amount of neighbours, for which type of tumour is it classifying? Is it just measuring the general accuracy, or is measuring the accuracy of classifying a specific type of tumour? Which one would it be, malignant or benign? In addition, if it was malignant, for example, is there a way to change the code so that it would be measuring the accuracy for a benign tumour instead?
Forgive me if it's a bit too confusing to read.
Forgive me if it's a bit too confusing to read.
-
bfinio
- Expert
- Posts: 761
- Joined: Mon Aug 12, 2013 2:41 pm
- Occupation: Science Buddies Staff
- Project Question: Expert
- Project Due Date: n/a
- Project Status: Not applicable
Re: Using the Breast Cancer Diagnosis Project for a Science Fair
Here is a reply to your question from the author of the project:
Accuracy is the measure of the correctness of a model's predictions, so it is measuring how correct its predictions are. For instance:
If the data points were as so: Malignant, Benign, Benign, Malignant,
and the model predicted it as so: Malignant, Benign, Benign, Malignant,
then the model would have an accuracy of 1, or 100%.
If the data points were like this: Benign, Benign, Malignant, Benign
and the model predicted this: Benign, Malignant, Malignant, Benign,
then the model would have an accuracy of 0.75, or 75%.
The model isn't classifying for just one type of tumor, it is given a tumor and then classifies it as Malignant or Benign.
Accuracy is the measure of the correctness of a model's predictions, so it is measuring how correct its predictions are. For instance:
If the data points were as so: Malignant, Benign, Benign, Malignant,
and the model predicted it as so: Malignant, Benign, Benign, Malignant,
then the model would have an accuracy of 1, or 100%.
If the data points were like this: Benign, Benign, Malignant, Benign
and the model predicted this: Benign, Malignant, Malignant, Benign,
then the model would have an accuracy of 0.75, or 75%.
The model isn't classifying for just one type of tumor, it is given a tumor and then classifies it as Malignant or Benign.