Research papers on Classification in Machine Learning
Classification is the task to assign probability to different categories based on how an object (like image and text) is likely to match the category. Intense research has been done in this domain. We have presented Research Papers in this domain using Machine Learning approaches.
Following are the research papers that you must go through to understand Classification in Machine Learning:
- Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review (by MIT)
- Very Deep Convolutional Networks for Text Classification (by Facebook Research)
- Accelerating Very Deep Convolutional Networks for Classification and Detection (by Microsoft Research)
- Active Convolution: Learning the Shape of Convolution for Image Classification
- Random Subwindows for Robust Image Classification
- CNN-RNN: A Unified Framework for Multi-label Image Classification
- Randomized Clustering Forests for Image Classification
- PCANet: A Simple Deep Learning Baseline for Image Classification?
- Are Sparse Representations Really Relevant for Image Classification?
We will go through the papers in short and share the links to access it accordingly:
Deep Convolutional Neural Networks for Image Classification: A Comprehensive Review
This paper summarizes the various CNN approaches towards Image Classification. This is a good starting point to get a good understanding of the overall picture. This will help you in understanding the other insightful papers following this.
Authors include Waseem Rawat and Zenghui Wang from MIT.
As of July 2020, it has over 645 citations.
Learn more by going through the paper on MIT Press Journal (HERE)
Very Deep Convolutional Networks for Text Classification
This is a must read as this has been a significant paper from Facebook AI Research in this field of Text Classification. It uses Machine Learning ideas. It introduces a new model VD-CNN which performs better than other existing models like RNN, LSTM and CNN.
As of July 2020, it has over 517 citations.
Authors include Alexis Conneau, Holger Schwenk, Yann Le Cun (all 3 associated with Facebook AI Research) and Loic Barrault associated with University of Le Mans, France.
Learn more by going through the paper on ArXiv (PDF)
Accelerating Very Deep Convolutional Networks for Classification and Detection
This is a must read as it improves over the previous paper by Facebook AI Research to improve the performance. This comes from Microsoft Research. It brings in nearly 4 times speedup with a small 0.3% increase in top 5 error.
Authors include Xiangyu Zhang, Jianhua Zou, Kaiming He and Jian Sun and are associated with Xian Jiaotong University, Xian, China and Microsoft Research, Beijing, China.
As of July 2020, it has over 365 citations.
Learn more by going through the paper on ArXiv (PDF)
Active Convolution: Learning the Shape of Convolution for Image Classification
This is an insightful paper. It introduces a new Convolution unit and addresses key disadvantages of Convolution. The new unit has no fixed shape as it is learned during backpropagation.
Authors include Yunho Jeon and Junmo Kim from KAIST.
As of July 2020, it has over 93 citations.
Learn more by going through the paper on CVPR (PDF)
Random Subwindows for Robust Image Classification
This presents a novel approach over ensembles of extremely randomized decision trees. This has been published at International Conference on Computer Vision and Pattern Recognition (CVPR).
Authors include Raphael Maree, Pierre Geurts, Justus Piater and Louis Wehenkel and is affiliated with Institut Montefiore, University of Liege.
Learn more by going through the paper on "University of Liege" site (PDF)
CNN-RNN: A Unified Framework for Multi-label Image Classification
This has been collaboration between Baidu Research, Facebook Speech, University of California, LA and Horizon Robotics.
As of July 2020, it has over 555 citations.
Learn more by going through the paper on CV Foundation (PDF)
Randomized Clustering Forests for Image Classification
Authors include Frank Moosmann, Eric Nowak and Frederic Jurie.
As of July 2020, it has over 387 citations.
Learn more by going through the paper on IEEE (PDF)
PCANet: A Simple Deep Learning Baseline for Image Classification?
PCANet is a basic Machine Learning approach that uses basic ideas like Principal Component Analysis, Binary Hashing, Blockwise Histogram and others and builds over them.
Authors include Tsung-Han Chan, Kui Jia, Shenghua Gao, Jiwen Lu, Zinan Zeng and Yi Ma.
As of July 2020, it has over 940 citations.
Learn more by going through the paper on IEEE (PDF)
Are Sparse Representations Really Relevant for Image Classification?
This is an insightful research paper and is a must read. The impact of sparse data is still a research question and this paper sets your mind in a good position to think independently in this line.
Authors include Roberto Rigamonti, Matthew A. Brown, Vincent Lepetit with association from EPFL.
As of July 2020, it has over 243 citations.
Learn more by going through the paper on EPFL (PDF)
With this article at OpenGenus, you must have a good understanding of which Research Papers you must go through in the domain of Classification using Machine Learning approach. Enjoy.