One of the most challenging aspects of research is getting stuck in that dark territory of unconfidence in your project, your progress, the very potential of what you are seeking to discover. By nature, my project this summer is very independent. Today was a big day for me, in that I succeeded in a lot of areas where I had been struggling in my code, and I received thorough feedback (both critical and encouraging) to help refocus my direction in the coming month of work.

First of all, in terms of my project itself, things were a little shaky. If you recall from a previous entry, I switched projects in week 3 from Audio Processing to Backward Stream Processing. I am still glad I made the switch, as I would be in much worse shape if I was still wrangling with the audio processing software. I absolutely love my new project, though the past few weeks have not been without their frustrations.

I have learned that I am prone to making stupid mistakes, often minor, though sometimes they have accounted for a lot of wasted productivity. I have learned that I function best when I spend a few minutes to meditate, and then generate concrete goals and subgoals. I have discovered that it is easy for me to lose focus when I lose sight of my overall research goals.

As I have been told by a few graduate students who are no longer taking classes, the lack of direct feedback can be disconcerting. In undergraduate course work, you get feedback throughout the term via projects and exams, even regular homework assignments. There is a metric for performance, and results of merit have certain requirements (you can only do so well if you bomb your midterm, or getting an A on the big project of the class gives you a decent buffer for doing decently well). When you start doing research, that steady feedback disappears. The metrics became my own to create this summer. How am I doing? Is my direction successful? Am I using my time and resources well? What outside research is needed to understand this better?

For those unfamiliar, I am working with data in the RFID domain, though I will soon be experimenting with audio data (podcasts, for example). The focus of my research is Markovian Streams: imprecise, ordered data generated from probabilistic inference. What does this mean to you? You can use these streams to predict location over time (for somebody carrying an RFID-tracked device), or to find the best guess for textual representation of a snippet of audio. My contribution is the processing of streams in a backwards way. This is not as simple as turning the arrows around and analyzing the probabilistic stream in the other direction. The basic idea is that this data generates probabilities at different time steps of what is happening (whether or not you are in your room or outdoors, or if a word in a podcast is ate or eat, etc). Sometimes it might be more efficient to process these streams backward, rather than forward in time.

In any case, today was a really big day for me. I fully implemented my steam reversal algorithm, and am pretty much ready for testing, which is really, really cool because my contribution to the Lahar (the data processing system I am working with) seems to work. I am very nearly done with the forward processing adjustments I need to make, and am really excited to see where this goes. Most importantly, I got a very concrete affirmation from Julie (my grad student advisor) that I am on the right track, that my progress is solid, that I can achieve this and apply the work I have done to audio data.

I have also second guessed my decision to go into research a few times this summer. Admittedly, a factor of my doubt was the money that I could make in industry. Some industry projects are really quite cool. At the same time, research is what has made me happy since I wandered into the Sophie LeLievre lab at Purdue university over 5 years ago, bright-eyed and science fair bound. Yes, sometimes stuff just does not work. Sometimes goals that you work toward for months yield satistically insignificant results. In many cases, the greatest challenge is to excercise the mind.

I realized today how much I absolutely love and desire a life of scientific research. This is the challenge, the competition, the lifestyle, and the passion that I am cut out for. I love the sensation of striding into unexplored territory, of developing something new, of the pursuit of knowledge in the name of scientific discovery. Even if I do not end up in the exact field of CS that I am working in this summer, I am confident that I will be very happy in graduate school.

Also, I wear glasses with thick black frames. They are pretty nerdy, I thought I should note that. Finally, there are some days when I just say Thank you, Science, for inspiring me today. Today is one of those days. Thanks, Science!