Sihan Chen Email & Phone Number

New York, New York, United States

- Led the full-cycle development of automated risk recognition, management, and interpretability services at TikTok Shop Mall, driving cross-functional team collaboration to enhance platform governance and customer experience.- Developed and deployed large-scale distributed machine learning systems to effectively govern product quality and combat negative actors on the platform.- Leveraged cutting-edge deep learning technologies to significantly solve cold-start problems for new markets and reduce negative reviews by optimizing data architecture and model development processes.

Mountain View, California, United States

- Developed and implemented advanced machine learning models for early detection of high-risk advertisers, markedly contributing to the platform’s safety and operational efficiency.- Architected a high-performance, real-time model serving infrastructure, ensuring rapid response to potential threats and maintaining high decision accuracy.- Innovated in model interpretability, drastically reducing manual review workload and improving decision transparency for advertising suspensions.

Mountain View, California, United States

San Francisco, California, United States

1. Image Segmentation: Built a 3D Residual U-net to precisely segment metastases from brain MRI images with PyTorch. Evaluated the effects of number, size, and locations of metastases on the accuracy, which has resulted in a scientific conference presentation and a manuscript. The result helps design a state-of-art model.2. Federated Learning: Simulated federated learning for brain metastases segmentation tasks, achieved close results to when training locally. This technique enables the lab to collaboratively learn a shared prediction model while keeping all training data on original hospitals in the future.3. Generative Modeling: Using CycleGAN and Pix2Pix to translate DRR Images to X-ray images.

Wuhan, Hubei, China

1. Analytically Solved the physics model of slim accretion disks and derived the ODEs.2. Simulated the accretion process of slim accretion disks around supermassive blackholes using Scipy, computed 10+ physical properties of the disk and derived the radiation spectrum. Applied the results to tidal disruption events, which was chosen as an outstanding thesis of the year at the school of physics.

The Kingdom Of Bahrain

1. Data Scraping: Built a web-crawler using Scrapy, scraped 10000+ ISO standards for different industries from the ISO official website into the company’s database. 2. Data Processing: Built a Python pipeline with Pandas, automated the excel spreadsheets tasks, which lower the time cost of managing the work progress of employees (saved 5 hours every week).3. Data Visualization: Analyzed data of the tourism industry of Bahrain, resulted in an industry analysis report.4. Built an ERP in Excel by using VBA.

Master'S Degree, Data Science

Relevant courses: 1. Relational Database 2. Machine Learning 3. Deep Learning 4. Distributed Computing 5. A/B testing 6. Data Structure.

Bachelor Of Science - Bs, Physics

Activities and Societies: Chairman of the Publicity Department in the Student Union of Physics SchoolBuilt mathematical foundations.

Visiting Student, Sociology And Psychology

Activities and Societies: T.i.n.Y Chinese Theater GroupDid case studies of organizations and society, which inspired me to apply data.