標籤:

Top 機器學習會議

Top Conferences:

ICML

KDD

NIPS

In my opinion, these three are the flagship machine learning conferences. They are the largest by attendance, attract researchers from across virtually all areas of machine learning, and have high visibility in industry and other computational fields.

Compared to ICML and NIPS, KDD is a bit more focused on new applications and less focused on basic methodology -- but many people consider KDD to be the more well-rounded machine learning conference. And remember, before there was Kaggle, there was the KDD Cup.

Smaller Conferences:

AISTATS

UAI

These two conferences typically span a wide range of topics in machine learning, although not quite as wide as the aforementioned three. They are also significantly smaller than the top 3, which makes them less visible to researchers outside the machine learning community. However, in terms of dissemination within the machine learning community, these conferences are just as good as the top 3. For instance, I regularly check up on papers coming from these conferences. But I would never physically attend these conferences if I didnt have a paper to present -- theyre not the best networking events because of their limited scale.

Niche Conferences:

ICLR

COLT

I call these two conferences niche conferences because they focus on a very narrow set of topics (from a machine learning perspective). ICLR is a recently created conference organized by the deep learning folks. The focus of ICLR is to study how to learn representations of data, which is basically what deep learning does. COLT is the conference on learning theory, and so is primarily focused on theoretical aspects of machine learning. Both conferences are great for their respective topics, and you get a more focused audience for your work.

Regional Conferences:

ECML

ACML

There are some regional conferences as well. Id attend ECML (resp. ACML) if I wanted to network with Europeans (resp. Asians).

Other Conferences:

Many conferences of other fields have machine learning papers or even a machine learning track. For instance, conferences focusing on vision, natural language processing, or information retrieval have the majority of their papers using machine learning in some fashion, and also have many papers that propose new machine learning techniques.

Hence, I often skim through the proceedings of the following conferences from other computational fields:

CVPR

ICCV

ECCV

SIGIR

CIKM

WSDM

ACL

EMNLP

SDM

ICDM

WWW

Finally, there are conferences that are so broad that you could even call them unfocused. But they do have a fair amount of machine learning papers.

AAAI

IJCAI

Journals:

The two main machine learning journals are Machine Learning and JMLR. Both contain top quality content. Other journals that are broader than machine learning are TKDE and JAIR. Both also contain some great machine learning papers as well.

Note that JMLR and JAIR are completely open access, so they are free to browse. TKDE and Machine Learning are behind paywalls, however authors retain certain copyrights so you can usually find the papers on Google Scholar or the authors home pages.

19.9k Views · View Upvotes


推薦閱讀:

Tensorflow VS PMML
1-2 Welcome
IBM Watson首席技術官:機器學習的三個挑戰
神經網路的理解
Hidden Markov Model(隱馬爾可夫模型(Discrete))

TAG:機器學習 |