You can use the widget Neural Networks. Yo can find an example recognising digits trained with the MNIST database in the book "Data Science and Engineering - Volume 3 Advanced Learning, Deployment and Deep Learning: with exercises in Orange, Python and Keras/Tensorflow"
How can I test time series; we can't use random testing in cross validation or random sampling . Is there a way to test based on time so test only the previous data which is before given time ? Thank you for your help
When I open up my Test & Score or Test Learners widget it shows me the results as Regression results (such as MSE, MAE & R2). How can I change it so it shows me Classification results (like the ones you see when you use the Test Learner, CA, Sens, Specs, etc)?
Sadly, you still need to program if your data isn't formatted properly -- it's extremely difficult to transform a mix-pivot data into something orange can use.... Also - if you want to calculate a standard low-level statistical test, like a mean-deviation - you are going to be doing custom python scripting.
i want to find sentiments in text file using orange and classify by using SVM but it will generate error target variable set ho to set target variable using text file in svm
Nice tutorial. But how do you find this much classifier (neural network,cn2) in Orange. I use orange 3 but i dont have these classifire. I only can use these classifire when i call orange from python. Can you tell me how you can add them in the gui?
Hi, I am using Oranges software v.3.22. How can I display the result of specificity and sensitivity? Because in your version in the video, it shows under the performance scores.