3 ms·
Are you planning to reproduce the experiments in every paper too?
by neur0mancer 13y ago
Are you planning to reproduce the experiments in every paper too?
- Swizec 13y agoI do not. While I would really like to be able to do that, the reality is that I just can't. I mean, I have a biologist friend who regularly clones bacteria and implements binary functions in their DNA. Not only do I not know how to hold a pipete, I just don't have the equipment to perform experiments like that. The goal is to get a functional understanding of as many different topics as possible. Enough to explain things in layman's terms, but I could not possibly become an expert in everything.
- neur0mancer 13y agoAt least for some of the Machine Learning and related papers (specially old ones), to reproduce the experiment should be relatively easy.
- TeMPOraL 13y agoReproducing those would also be a big step towards understanding them better.
- neur0mancer 13y agoAnd also a big step to catch these papers in which the results are not exactly as good as the authors said..
- StefanPatelski 13y agoEven if he would work full-time on that, it would be impossible to do in most cases. A week is usually not enough time to organize an experiment, especially when you are not experienced in the field.