Fei Liang Email & Phone Number

Ithaca, Ny, Us

Ithaca, Ny, Us

Organize regular graduate student activities to help students get familiar with each other and relax from courseworkParticipate in the Master of Engineering Exposition to guide prospective students in getting familiar with our program

New Brunswick, Nj, Us

Cornell M.Eng Data Science Capstone Project

北京, Beijing, Cn

Designed a nationwide air quality forecasting platform to provide real-time pollutant concentration prediction and visualization to guide industrial and academic decisionsManaged more than 21,000,000 pieces of data from 2396 meteorological stations to excavate patterns using Python libraries such as PyTorch, Pandas, NumPy, etc.Built a time-space model by combining Graphical Convolution Network and Long-Short Term Memory model to achieve an over 95% pollutant level predictionDesigned interactive heat maps, line charts, boxplots and other graphical data visualization methods using Python libraries such as folium and pyecharts, and embedded them into the air quality forecasting platform

Designed supporting models for the online intelligent transportation platform to provide gas station suggestions for over 5,000,000 truck drivers and 400,000 transportation companies.Upgraded gas station classification algorithms by applying the DBSCAN algorithm on more than 50,000 truck stops Formulated practical insights by extracting, manipulating, analyzing large volumes of data using Hadoop and Python libraries such as Pandas, NumPy, etc.Collaborated with Software Engineering team to evaluate model performance and improve model accuracy.

Generated a cardiac MRI image segmentation system based on images form 70+ patients to facilitate doctors diagnosis. Implemented a U-Net model to achieved 71% accuracy on segmenting tissue scars by data normalization, hyperparameter tuning, and data augmentation. Directed a team of five people to formulate solutions to technical questions proposed by specialists of the hospital.

University Park, Pa, Us

Implemented a deep learning-based algorithm to numerically solve partial differential equations which are used to describe heat transmission, wave propagation, Alzheimer disease development, etc.Designed a training algorithm by combining random points selection with Stochastic Gradient Descent to increase the training five times faster than the traditional method for 2-D modelGenerated results and errors plots using Matplotlib to demonstrate result and error distribution for further analysis

Founded the Computer Vision Group and won 3rd place in the 2019 international Robomaster CompetitionOrganized weekly group workshops and more than 10 extracurricular activities to expand the size of the club from the original 9 members to over 30 members

Master Of Engineering - Meng, Operations Research

Bachelor Of Science - Bs, Computational Mathematics

Bachelor Of Engineering - Be, Electrical And Electronics Engineering