What is the AutoML?

• A trial-and-error process means that even experts require substantial resources and time to create well-performing models. To solve repetitive and experimental learning, automated machine learning (AutoML) is a promising solution for building a DL system without human assistance.

  

• Selected the review paper on the state of the art in AutoML field, published in 2021

  • Paper: https://arxiv.org/pdf/1908.00709v6.pdf

  • It covers the 4 categories such as data preparation, feature engineering, model generation and model estimation

    

Data Preparation

• IMAGE
• Consists of 3 major categories, data collection, data cleaning, and data augmentation

Data collection

Data Searching

1. J. Yang, X. Sun, Y.-K. Lai, L. Zheng, M.-M. Cheng, Recognition from web data: a progressive filtering approach, IEEE Transactions on Image Processing 27 (11) (2018) 5303–5315.
   • Paper: Yang et al. [51] proposed an iterative algorithm for model training and Web data filtering.
   • URL : https://xiaoxiaosun.com/docs/2018-tip-Recognition-from-Web-Data.pdf

Data Synthesis

1. I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. WardeFarley, S. Ozair, A. C. Courville, Y. Bengio, Generative adversarial nets, in: Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, K. Q. Weinberger (Eds.), Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, December 8-13 2014, Montreal, Quebec, Canada, 2014, pp. 2672–2680.
   • Paper: deriving synthetic data is Generative Adversarial Networks (GANs) [71], which can be used to generate images [71, 72, 73, 74], tabular [75, 76], and text [77] data
   • URL : https://arxiv.org/abs/1406.2661

Data Cleaning

Fixed Dataset