Presumably the goal is to have a lot of chips at the end (more than you started with?). I would assume if you bet 10 chips, you get 10 for being right, and zero for not. So betting 0 chips all the time probably won't get you the job!
Can't you? I have just been using the lightning class as a wrapper, with my normal model on self.model or something like this. Same for the dataloaders.
I like the idea of kubernetes, and (some time ago) worked through a basic tutorial. My main confusion is how to set up a development environment. Are there any guides you could suggest that cover basic workflows?
The keyword to search for, if you are interested, is epigenetics. Modification to e.g. histone protiens can change the expression of genes through a variety of mechanisms. As far as I'm aware, exactly how this works isn't that well understood.
I had a macro in emacs that wrapped the highlighted text in some xml tag (say <book></book>). Processing that I could label it however - e.g. IOB or whatever you fancy. The labelling didn't really take that long to do, maybe a few hours over a couple of days.
I did this a while ago (www.hnreads.com, bookbot.io). Manually labelled ~2k comments, trained a NER system, then validated the titles via amazons api. Was pretty easy once I had manually labelled comments. Don't recall my F1/precision/recall scores - they were ok but lower than the state of the art reported in papers.
We just had a paper accepted after 11 months. There is no good reason for it taking this long. I wish biology would adopt the publishing practices of physics or machine learning.
I've moved completely from Vim to Spacemacs. I can't see myself going back to plain Vim any time soon. The emulation via evil is very good, and the layers provided are generally immediately usable with consistent keybindings.
I haven't regretted doing my PhD in the slightest. I've learned a huge amount, in a wide variety of areas. However, the system in the UK is a bit more favourable regarding funding. Having said that, the funding doesn't scale with location. I live in one of the most expensive cities in the UK (outside London), and am funded to the same level as everyone else outside of London.
The bigger problem I have is in deciding to leave academia. There are still questions I would like to know the answer to (and I expect there always would be). It just doesn't seem to make sense to take that path. Low pay, high competition, and the likelihood of being on a series of 18month contracts for the majority of your career. Indeed, in my specific field (Bioinformatics) the career path isn't even that well defined, and you can more easily get a job at a software company who will pay several multiple of what you could expect as a postdoc. It just doesn't make sense to do it.
I like this idea! In my research we deal with relatively large amounts of sequence data, all of which needs to be associated with geo (https://www.ncbi.nlm.nih.gov/gds/). While geo is in many ways a good thing, it is not the most pleasant to use - I would love it if we could use something like torrents instead.
I feel like there is a danger, however, that using torrents would facilitate the thousands of nonstandard (often redundant) formats bioinformaticians seem to create.
I made something last winter (http://www.hnreads.com/), though I haven't fixed a couple of bugs or updated it since. I used named entity recognition rather than links to find book titles, perhaps something you might consider if you wanted to expand the site. I like the idea of showing the top books per time period!