It’s already October (!) and the fact that I’ve managed to forget writing anything in the progress reports category, probably tells you something about my progress… Nevertheless, I’m holding myself accountable, so here a few short updates.
I finished THE survey on semi-supervised, multiple instance and transfer learning I’ve been working on in April and uploaded as a preprint to arXiV, inviting feedback from readers. This was quite scary to do, but in the end also very rewarding. I now submitted the paper to a journal.
Progress is slower than I would have hoped overall – to catch up with my goal I would need to finish two other papers by the end of the year.
A strategy that is helping somewhat is to aim to write 250 words a day. That’s not a lot of words, but I notice that on the days that I do it in the morning, I am probably going to write more than that, whereas on the days that I ignore it, I will not end up writing at all. Having a Kanban board in my office where I see in which stage each paper is, also helps – thanks to Eiko Fried for the idea.
Once I’ve relaxed my blogging once a week goal, I’ve started slacking off on this… It has been good for me to not have this obligation on top of everything else, but at the same time it is a shame because in the end I feel like this is one of the most important things I’m doing. What was great is that at one point I prescheduled posts for several weeks. I’d like to do this again, but it’s difficult to get into a mindset of writing a lot upfront.
I did get a little bit of income (perhaps $25) from the blog, especially after the How I Fail post by Ian Goodfellow went a bit viral. But because my provider was so unhelpful about the problems the traffic caused, I also had to switch providers, so it’s definitely not a profitable business 🙂