陈力'blogs

Resume

Introduction

Name: Chen, Li
Date of Birth: 10-10-1992 
E-mail: vchenli@csu.edu.cn 
Website: http://www.vchenli.com 
Research: machine learning, deep learning, interpretable AI systems

Education

Date: 09,2018 – now  
University: Central South University  
Degree: Ph.D. candidate
College: School of Geosciences and Info-Physics
Major: Geomatics Science and Technology
 
Date: 09,2015 – 06,2018 
University: Central South University  
Degree: Master (top 5%, GPA 3.7)
College: School of Software     
Major: Software Engineering 

Date: 09,2011 – 06,2015
University: Central South University
Degree: Undergraduate (top 5%, GPA 4.0)
College: School of Software
Major: Software Engineering

Publications

  • Li H.F., Chen L, Ding H.L., et al. Procedural Learning with Robust Visual Features via Low Rank Prior[J]. IEEE Access, 2019, doi:10.1109/ACCESS.2019.2894841. [In press] paper
  • Li C, Hailun D, Qi L, et al. Understanding the Importance of Single Directions via Representative Substitution[C]// AAAI Workshops. 2019. paper
  • Chen L, Ding H, Li Q, et al. Adversarial Feature Genome: a Data Driven Adversarial Examples Recognition Method[J]. arXiv preprint arXiv:1812.10085, 2018. paper
  • Hongxiao F, Li C, Haifeng L, et al. Information Geometric Measurement of Internal Transfer of Deep Neural Network[J] Journal of Hunan University of Science and Engineering; [In press] paper
  • Chen L, Fei H, Xiao Y, et al. Why batch normalization works? a buckling perspective[C]//Information and Automation (ICIA), 2017 IEEE International Conference on. IEEE, 2017: 1184-1189. paper
  • Hongxiao F, Li C, Jiabao H, et al. SANet: An Approach for Prediction in Music Trends[M]//Advances in Computer and Computational Sciences. Springer, Singapore, 2017: 383-389. paper
  • Chen L, Fei H, Ding H, et al. A data sample method based on double decision tree[J]. Computer Engineering & Science, 2019, 41(01): 130-135. paper

Honor Reward

  • 2018 Excellent Graduates of Hunan Province
  • 2017 China Graduate Mathematical Modeling Competition, third prize
  • 2015 China University Student Service Outsourcing Innovation and Entrepreneurship Competition, second prize
  • 2015 national scholarship for graduate students
  • 2015 Excellent Graduates of Hunan Province

Project / Internship Experience

Date: 11,2017 – now
Subject: Remote sensing image recognition and interpretable AI systems

  • The National Natural Science Foundation project. Aiming at open scientific problems of image automatic recognition and understanding of the computer vision, we want to solve the problem of invariance of the visual representation, the invariance feature of deep neural network internal learning process. This study provides a theoretical basis for further exploration of the underlying mechanism behind the vision, and provides new and potential ideas for machine vision in the era of big data.
  • As the leader of the sub-project team, I developed methods to measure the deep neural network changes in the learning process, and implemented the corresponding technical solutions. Additionally, we use the methods of visualization and other methods to analyze the interpretability of deep learning. We try to apply it to the problem of the adversarial example, and obtain a good recognition result.

Date: 06,2017 – 09,2017
Subject: Ship detection in complex scenes

  • The program realizes multi-scene and multi-scale detection and identification of ships, and achieves the detection task of small targets.
  • Responsible for model implementation and programming. According to different scenarios and situations, data preprocessing is performed and various types of U-NET networks are implemented.

Date: 06,2016 – 05,2017
Subject: Change measurement of transfer learning in deep neural networks in information geometry

  • I have discussed and studied with many teachers and proposed solutions for the project. This paper mainly discusses the black-box problem of deep learning through information geometry method, and tries to solve the stability of representation of deep neural network.
  • I learned about machine learning and deep learning. The ability to reproduce the paper algorithm has been improved.

Date: 05,2015 – 06,2016
Subject: Analysis of temporal and spatial characteristics of traffic flow micro-structure, project leader

  • The school research project. In response to the increasingly serious traffic congestion problem, we use the subspace learning method to find the “network traffic flow” from the macro traffic flow based on the analysis of traffic data and behavior data. The “feature subspace” is a distribution function that expresses the temporal and spatial characteristics of the city.
  • The project won the excellent project award, and the related achievements also won the second prize of the national service outsourcing competition.

Date: 06,2014 - 11,2014
Subject: Teradata company internship, Data Analyst

  • The project allowed me to go through the entire data processing and improve my understanding of the data. I mainly implemented some data preprocessing methods and learned some machine learning algorithms.

标签:

发表于2011-09-03 11:51:24,最后修改于2019-07-17 16:18:56。

本站文章欢迎链接分享,禁止全文转载。


下一篇 » 个人简历

推荐阅读

Big Image