Ao Shen
+1-447-***-**** ć ad1s7p@r.postjobfree.com ] Ao Shen
EDUCATION
University of Illinois Urbana-Champaign Aug 2023 - Dec 2024 (Expected) Master of Computer Science Urbana, IL, USA
• Member, Graduate Chapter of Women in Computer Science (GradWCS) Hefei University of Technology Sep 2019 - Jun 2023 B. Eng. in Computer Science and Technology (Innovation Experimental Class) Hefei, China
• GPA: 93.85/100, Rank: 1/179, 2022 National Scholarship (Top 0.2% national-wide) SKILLS
• Programming: Python, Java, C/C++, JavaScript, TypeScript, SQL, HTML, CSS
• Web Development: React, Vue3, MyBatis, Node.js, Express.js, Spring-boot, Django, Ant Design
• AI/Data: PyTorch, NumPy, Pandas, Matplotlib, Hadoop, Spark WORK EXPERIENCE
Shanghai Artificial Intelligence Research Institute Co., Ltd. Oct 2022 - Aug 2023 Machine Learning & Algorithms Engineer Intern(Technology Innovation Department) Shanghai, China
• Developed a time-series prediction platform with Django and React, integrating advanced algorithms like LSTM-CNN hybrid and Transformer-based TFT model, achieving a 13% increase in forecast accuracy.
• Created a new unsupervised abstract classification method, based on DocSCAN and a ChatGPT-generated pseudo- dataset that was only 1% the size of PubMed 200k RCT, achieved an accuracy within 10% of supervised methods.
• Synthesized two Chinese medical abstract classification datasets using GPT-3.5 and achieved nearly 90% accuracy clustering 100 of the synthesized data, which is comparable to a real PubMed dataset with true labels. YiheCode Technology Co., Ltd. Dec 2021 - Jun 2022
Machine Learning Engineering Intern(AI Product Department) Beijing, China
• Developed a Business Intelligence SystemwithSpringBootandReact,serving both external customers and supporting internal teams with automated analysis and visualization tools.
• Led the creation of a responsive frontend using Ant Design Pro and TypeScript, automated requests with Umi Ope- nAPI, and enhanced data visualization capabilities with Echarts.
• Integrated Redisson’s RateLimiter for effective rate limiting, achieving sub-200ms response times and over 5,000 RPS through asynchronous AIGC request processing with custom thread pools.
• Implemented data sharding with MyBatis, message persistence with RabbitMQ, resulting in up to 14% improvement in query performance, and auto-generated API documentation using Knife4j and Swagger. PROJECT EXPERIENCE
News Text Classification System Jun 2022 - Oct 2022
• Developed a text classification system using machine learning algorithms, achieving benchmark accuracies.
• Enhanced model performance by designing and implementing a suite of deep learning models using PyTorch, includ- ing TextCNN, TextRNN, FastText, TextRCNN, DPCNN, Transformer, Bert, ensemble models like Bert+DPCNN and Bert+TextRCNN, reaching peak accuracies of up to 94.91%.
• Deployed a Flask-based web interface for instant news categorization, leveraging the refined deep learning models. Electric Power Supervision Assistance System Based on the Vision Question Answering May 2021 - May 2023
• Developed a user-friendly platform allowing users to upload images of electrical equipment, pose inquiries, and receive prompt and accurate responses.
• Explored and integrated the BLIP model as a third strategic advancement, enhancing the algorithm’s capabilities.
• Incorporated a feature for the system to retain a detailed record of user interactions and queries. Railway Technical Department Shunting Safety Protection System May 2020 - May 2022
• Developed a system based on domain adaptive deep learning to provide a shunting signal recognition system solution.
• Built an incremental learning network applicable to domain adaptation problems to avoid catastrophic forgetting.
• Detected signal lamps and railway lines through the adaptive target detection approach based on feature difference penalty to increase the robustness of the model.