Azure Machine Learning in dynamics

Question

I just started to work with Azure Machine Learning using Microsoft Azure Machine Learning Studio.

Could you please advise the proper way of using the Classification Model so that the Model can analyze the window/interval with series of information before it reaches the target/class which should be predicted further on by Model?

Our problem could be solved only by analyzing all the previous information in it's dynamic/evolution (e.g. depending on the dynamical change of the Patient's medical results it could be found out some sickness including its current stage and the stage which we could expect in the medium term).

For example in an input file, we do not provide in each row together with the variables the target info, it is shown only at the row/moment when the situation is matured to reach such target.

If it is already available some materials/tutorials on this subject in Azure ML or somewhere else I would highly appreciate such info and links.

Answer 1

I have worked with predicting diabetes in Azure ML Studio using one of the saved data set. It predicts if one has diabetes or not based on the parameters. I have used Two-Class Logistic Regression Classification here. You can surely move ahead with modifying this experiment. You can try and add a column that has previous medical history. Also, you can try to predict the type of diabetes. With this base idea, you can predict many other things.

Otherwise, it becomes a recommender system. 

Here is the tutorial that has the experiment: click here