Hi ML community,
I’m nearing the end of my three-year PhD journey. Throughout this period, I’ve dedicated myself to producing a research paper annually, targeting top-tier conferences like ICML, ICLR, and NeurIPS. Despite my efforts and resubmissions, none of my papers made it through. As a result, my publication record consists solely of three manuscripts on arXiv.
My initial post-PhD ambition was to delve deeper into machine learning research at leading tech companies such as Facebook, Google, or Microsoft. However, my applications were turned down, primarily due to the lack of publications in prestigious conferences, which seems to be a crucial criterion for these roles.
Confronted with this setback and the pressing need to manage my finances, I shifted my focus to more traditional industry roles in consulting and finance. I’ve recently secured a position in quant finance, which, while exciting, means I won’t have the bandwidth to revisit and resubmit my research papers.
Reflecting on this journey, I sometimes feel disheartened, questioning the value of my PhD experience, especially when I consider my lack of published work in major machine learning conferences.
I see other PhD students in my field publish 2 papers per year in these top conferences which makes me wonder whether I am a failure? I’m open to any thoughts or advice on my situation.
I never understood the logic peer reviewed conferences. For example take 2 top conferences in any sub field of ML, now why are these 2 conferences top conferences? Do they become top conferences just by limiting the number of papers they accept? The people who review for these conferences are also the same so what makes them different. Also the capabilities of each reviewer are different so how are scores from each reviewer given the same weightage? A incompetent reviewer might give a bad paper very high score whereas a competent reviewer might give a good paper average score.