ml4bio-workshop icon indicating copy to clipboard operation
ml4bio-workshop copied to clipboard

Materials for a workshop introducing machine learning to biologists

Results 59 ml4bio-workshop issues
Sort by recently updated
recently updated
newest added

A short quiz assessing machine learning skills and concepts may be helpful for participants to identify whether they want to attend the workshop or are already familiar with the basics...

Related to issue #29. While the image has been scaled down so it is no longer confusing, it would be nice to instead show this kind of data normalization and...

We do not have an example dataset that shows why it can be misleading to focus on performance metrics that include true negatives (accuracy and ROC curve) when the classes...

We could add material about what happens when a proper machine learning pipeline is not followed. A participant raised the [Duke cancer genomics scandal](https://en.wikipedia.org/wiki/Anil_Potti), but I don't think that is...

We should update the performance guide to explain the precision-recall curve of a random classifier and how the area depends on the class imbalance. Copied from #14

scikit-learn has a [DummyClassifier](http://scikit-learn.org/stable/modules/generated/sklearn.dummy.DummyClassifier.html). Adding one of these could be a good way to teach about benchmarking against naive baselines and understanding how performance metrics depend on class label distributions.

I'm removing the inline comments from the software guide and moving them here instead: - How are ties broken if a user selects the best classifier based on a metric?...

There is potentially relevant material in this blog post about developing new lessons for the (Software and Data) Carpentries: https://carpentries.org/blog/2018/07/curriculum-vision/ They have a [Handbook on Curriculum Development](https://carpentries.github.io/curriculum-development/) and are partnering...

This training uses the Galaxy platform and has similar goals as us, but it isn't solely focused on machine learning https://www.sciencedirect.com/science/article/pii/S2405471218302308