dsanalytics

@markotoplak use case is serious research - very weak is not the same as off in science. 2 + 2 + 0.00001 is not the same as 4 in science....

@janezd I do not want to get into a debate, but I disagree. IMHO, regularization is the worst type of solution to overfitting. I guess this is another sklearn devs...

@markotoplak in case you are putting reg-off option for NNs too, please check https://github.com/biolab/orange3/issues/5818 Thanks.

@markotoplak @ajdapretnar can you open this now? I completely agree with Marko's reasoning: > I like how explicit that is, and might be useful when teaching classes. Orange does have...

@ajdapretnar @markotoplak I'm quite certain that you're very busy, but would you mind finishing this? It's been almost 6 months. Thank you.

@markotoplak @janezd @ajdapretnar Dear Marko, My humble contribution are suggestions/ideas and bug reports submitted over the years. Not big, but not nothing either. Orange team fixed some of those bugs...

@ajdapretnar @janezd I only cc-ed, so to speak, you and the professor to keep you in the loop - that's all. Regards

@janezd @ajdapretnar @markotoplak Thank you all - appreciated.

@MarcusTarcus You can use MWSnap (free) and set it up to autosave what you capture and prompt for a file name if you set it so or autonumber name as...

Quick article regarding this: https://www.johndcook.com/blog/2012/12/20/basics-of-sweave-and-pweave/