I'm a software development engineer at Amazon, working on Manga retail targeting Japan marketplace.
I'm a graduate student in Caltech major in Electrial Engineering, whose graduation time is December 2020. I got my bachelor's degree in Computer Science in Zhejiang University, China. I used to be a NLP algorithm intern in Alibaba, Ant Financial for three months. I'm responsible for a machine learning model aims at predicting users' default rate according to the title of goods in their purchase history. With the help of my model, the predition accuracy improved by 10% at the end of my internship.
During my undergraduate year, with Prof. Ming Chen being our supervisor, I used to take part in Team ZJU-China iGEM-2017 and won international gold medal as a team. Then I spent my junior working with Prof. Gang Pan at ZJU on a machine-learning-based video interpolation project, the paper of which was published on CPGIS 2018. With Prof. Quanzheng Li at Harvard Medical School I spent ten months on my final year project.
Worked on Customer Cohort project to allow manga promotion based on customer reading history.
Implemented GBDT Model to concatenated titles of all items from users’ purchase history as their document.
Built the user feature dataset by Chinese word splitting, vector space generation and feature discretization.
Improved the KS of default rate prediction from 0.33 to 0.44
Project 1: Accepted for presentation by ISBI 2019. Novel Radiomic Features Based on Graph Theory for PET Imaging. We proposed a series of new radiomic features for PET image analysis base on graph theory and network analysis. The evaluation result indicates that our graph-based features reveal significant and unique information of tumor in PET images.
Project 2: Accepted for presentation by ISBI 2019. We proposed a new CMR image-based framework to classify heart diseases with different etiologies. For leveraging more 3D structural information, the approach combining MPCA and LDA significantly improves classification accuracy. Based our method, the individual abnormal myocardium can be automatically detected and visualized.
We Used Gene-modified Trichoderma Endophyte to Assist Economic Plants in Defensing Common Diseases Represented by Blackspot to Improve Survival Ratio and Harvest.
Our project can generate the middle frame by analyzing an existing video by combining the advantages of estimating optical flow and hallucinating the RGB value directly, which is able to generate sharp and realistic frames
Majoring in Electrical Engineering.
GPA: 3.90/4.00
Ranking: Top 5% out of 208
Majoring in Computer Science and Technology, estimated to obtain a bachelor's degree in Jun 2019.
Z. Zhou, N. Guo, J. Cui, X. Meng, Y. Hu, H. Bao, Q. Li. Novel Radiomic Features Based on Graph Theory for PET Imaging. IEEE International Symposium on Biomedical Imaging
S. Qian, Z. Zhou, S. Lai. Frame Interpolation Using Deep Pyramid Flow. The International Association of Chinese Professionals in Geographic Information Science
H. Bao, H. Ren, Z. Zhou, X. Li, N. Guo, Q. li. 3D Regional Shape Analysis of Left Ventricle Using MR Images: Abnormal Myocardium Detection and Classification. IEEE International Symposium on Biomedical Imaging
X. Zhao, X. Li, N. Guo, Z, Zhou, X. Meng, Q. Li. Multi-size Computer-aided Diagnosis of Positron Emission Tomography Images Using Graph Convolutional Networkds