I am really lacking a cool idea for project that can be solved by applying Statistical Pattern Recognition - supervised learning with class labels (no clustering). If you have any ideas I'd be happy if you could share them with me. I am looking for something that can be done within ~ 2-4 months. Right now, I am passionately studying Statistical Pattern Recognition and I am looking for some application that really excites me. Ideally, something useful of course, because I'd be happy to share it with other people as an open source project on GitHub in the end. I tried brain storming, but all I could came up with was old hat - something that has been solved before... Protein fold prediction, protein secondary structure prediction, handwriting recognition. Also "Google" hasn't provided me with anything useful...
I am looking forward to your suggestions!
About my background: I am working in the field of Biochemistry w. focus on Protein Science, so if you have any suggestions related to it (maybe something that you would find useful) this would be great. But also ideas from completely unrelated fields are very welcome!
[link][8 comments]