Its been a year since I finished my MS program. I had come to US with the intension of doing a PhD, but a couple of things did not work out for me while in grad school. Here are a few thoughts on grad school in general.
Reasons for ditching phd –
- Not a good fit with the lab and the topic I was involved in – The lab’s main focus was planning and scheduling. There was also work on the side about databases/information retrieval/social media analysis etc… I was involved in topic modeling/social media analysis, and very soon lost interest in it. I was interested in graphical models(non-parametrics) and online learning in ML( the sort of stuff Zoubin Ghahramani and Shai Shalev-Shwartz work on). I thought I could make a bargain and settle for planning/scheduling etc… but I could not see how I could meaningfully change tracks to planning, so I bailed out.
Nonetheless, there are a lot of interesting things I take away from my MS experience. Here are a few non-technical ones –
- Do a Masters thesis only if you are seriously considering doing a PhD. A MS thesis will give you a ringside view of what a PhD will be like. An MS thesis for any other reason does not make sense.( I already knew this getting into the program, but the number of people who did not know it was staggering, hence mentioning it here)
- Get your finances sorted out within the first semester of your MS program( Half a semester or One Semester fee is what people can usually pay without taking any sort of loan). Try to get fully funded. Now it is unfair to expect your professor to fund your research while you are still an MS student because it does not work out well for them. But there are usually a lot of web development and other software development work that gets offered on campus which will give you full assistantship. Try to get those. You cannot do good research if you have an education loan hanging on your head.
- I saw many people opting for PhD because it will give them more time to look for more lucrative jobs in Silicon Valley. Such people are going into a PhD program, confident that they will not finish the program and will quit in the middle. This is probably the lamest thing you can do.
- Also there are many people who went into a PhD program not knowing what exactly it entails. Their goal with a PhD program is to end up with a job in a top company like Google/Facebook etc… Now, Google is a brilliant company, and many brilliant PhD students find fulfilling work in Google, post their PhD studies. But getting into a PhD program with the aim of getting a particular job in a particular company is a very painful way of doing things.
- The motivation for a PhD should be learning as much as possible about a subject, getting excited about publishing your ideas, and building a network and making a name for yourself. PhD is not about taking an extra load of courses, and finding a high paying job.
- Try to align your research topic with the main focus of your research lab and your advisor. This will make your experience more fulfilling. Going off on a tangent can get you lost.
- If you follow up on your interests even after grad school, you can end up with an applied research position in some of the industry labs 4~5 years down the road even without a PhD. ( if that is your aim )
- Having a good rapport with your advisor or other senior PhD student/post-doc you are working with is most important. Without this everything else futile.
Few other points about choosing schools and professors –
- Choose school over professors. Better schools invariably have a good supply of sufficiently good profs, might not be a top prof but still it will give a lot of options if you want to switch specialization/prof etc..
- Also the top profs in mediocre schools are usually grumpy and unhappy that they do not have access to a better talent pool . Which makes them hard to work with, which in turn makes it all the more difficult for you. This happened a lot in the lab I was working in.
- I usually tend to work better when the subject is challenging and deep. If the task is easy/boring I tend to slack off and perform worse. I did not feel excited at the prospect of working in the lab I was involved in.
Other references related to grad education –
These stories are true in a weird way –