Getting starting of PhD

To start a PhD program, you must find someone to supervise you. I would not teach you how to find a supervisor as I am not well experienced in this. Usually, people will spend quite a long time without a referral. So, a referral is essential even in academics. In my case, I contacted my PhD supervisor through email, who is the connection to my MSc supervisor.

There is UK and US mode of PhD. Generally, the UK took three years, and the US took five years. Malaysia follows the UK mode, so I am expected to finish my PhD within three years. If I can do that, I am said to be graduated on time. Now, I will talk about the research proposal. You can look at my research proposal for admission here.

The proposal states the basis of what you want to research for the subsequent years. It is not expected to be too specific, and people understand you need time to refine it. Don't worry about the content or the title. It will change in most cases, mine too. In this proposal, I suggest using a graph neural network and active learning for semantic segmentation. Both are a topic in themselves, so I must choose one to continue for deeper exploration.

I spent a long time exploring a few topics that interest me: reinforcement learning, meta-learning, active learning, and graph machine learning. Looking at those topics, you know why I spent one year deciding on the topic to focus on. PhD is about depth, not breadth. I ended up with graph machine learning since I have been interested most since my MSc.

Next, we will look at exploring the topics to work on.