3 ms·
I'm a published researcher in AI. * Where: depends on the field of research. Top general AI conferences include AAAI and IJCAI, which occur once per year. Pape
by bo1024 9y ago
I'm a published researcher in AI.
* Where: depends on the field of research. Top general AI conferences include AAAI and IJCAI, which occur once per year. Papers are submitted online and reviewed anonymously. There are many others. Top machine learning conferences include NIPS and ICML; same comments apply.
* How: well, you write a paper conforming to the length and style requirements (this requires learning LaTeX, at least the basics). The paper should describe the problem you are trying to solve, briefly describe and cite relevant prior work and explain either why it does not solve the problem, or why the solution you will be presenting is better. It should then present your solution and evidence for it, be that theorems and proofs or experiments or simulations. You then submit the paper to a conference online.
You didn't ask how you can conduct your research on AI (your phrasing makes it sound already completed), so I'll try to avoid too much unsolicited advice, but this is the difficult part. It's important that you have read the most relevant papers to the problem you're trying to solve (this will help you see how to structure your own paper) and that you have a clear idea of where your research fits and improves on the existing literature.
Frankly, this will be very difficult unless you have some help from a mentor or collaborator who has been through the process before. Ideal would be to get in touch with someone who is doing research in the specific area you are interested in and ask to meet with them, explain what you are doing, and get their advice on related work, whether your work is publishable, and how to present it. That doesn't make them a co-author.
If you are interested in getting into research long term, ideal would be to ask a similar person to help or join on one of their projects, or collaborate on your idea. After some experience it will be much easier to figure things out for yourself.