李鹏宇
Machine Learning Engineer at tencent
Dongcheng District, Beijing, China
Summary
🤩
Rockstar🎓
Top School李鹏宇 is a Machine Learning Engineer with 11 years’ experience applying data mining, social network analysis, NLP, and deep learning to product and research problems. He has built data platforms and user profiling systems at Lenovo, analyzed user behavior and information propagation at Tencent, and provided data-mining consulting at SPSS China. His open-source contributions include improving training and prediction modules and adding an HMCN model to Tencent’s NeuralNLP-NeuralClassifier toolkit, reflecting practical expertise in neural hierarchical multi-label text classification. Holding a master’s in software engineering from the University of Chinese Academy of Sciences and a bachelor’s in applied mathematics, he bridges rigorous academic foundations with production-focused engineering. Based in Dongcheng District, Beijing, he combines hands-on model implementation with large-scale data engineering experience across Hadoop and Hive ecosystems. An understated strength is his knack for finding and fixing subtle training inconsistencies that improve model reliability in real-world deployments.
11 years of coding experience
3 years of employment as a software developer
硕士, 计算机软件工程, 硕士, 计算机软件工程 at University of Chinese Academy of SciencesBachelor, Applied Mathematics, Bachelor, Applied Mathematics at Linyi normal university
Github Skills (10)
neural-network10
pytorch10
machine-learning10
nlp10
trainings10
python10
text-classification10
modeling10
evaluation9
eval9
Programming languages (1)
Github contributions (5)
Tencent/NeuralNLP-NeuralClassifier
Jul 2019 - Apr 2021
An Open-source Neural Hierarchical Multi-label Text Classification Toolkit
Role in this project:
ML Engineer
ML EngineerContributions:18 commits, 15 PRs, 15 pushes in 1 year 8 months
Contributions summary:李鹏宇 primarily focused on improving the training and prediction modules of the neural text classification toolkit. They addressed inconsistencies in the loss function and training processes, ensuring correct model behavior. The user added a prediction module enabling the model to classify new text inputs. They also added the HMCN model, demonstrating their ability to integrate different neural network architectures into the project.
nlpdeep-learninghierarchicalmachine-learningmultilabel-classification
coderbyr/MV-LSTM
Aug 2017 - Apr 2018
Contributions:13 commits, 12 pushes, 1 branch in 7 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.