Back End Engineer

Jeffery Chen

*********@********.*** 669-***-**** www.linkedin.com/in/jefferychen9902

EDUCATION

University of California, Berkeley Berkeley, CA, USA Bachelor of Arts in Computer Science, GPA 3.915 Expected Graduation Date: 2022 Spring Completed Coursework:Data Structure, Efficient Algorithms, Probability and Random Process Current Coursework: Computer Security, Database Systems, iOS Development SKILLS

Technical Skills:Python, Java, C, SQL, Swift, PyTorch, Keras, Git, Linux Languages:English, Chinese EXPERIENCE

FunnelFoods Berkeley, CA, USA

Back End Engineer February, 2019 - October, 2019

● Developed in Python the core algorithm which recommends cooking recipes based on user’s ingredient storage.

● Improved automation process by engineering an NLP algorithm that corrects spelling from noisy OCR outputs.

● Collected and cleaned with NumPy multiple food-related datasets with thousands of entries. School of Electrical Engineering, Shanghai Jiao Tong University Shanghai, China Research Intern, EEG-based Emotion Recognition July, 2019 - August, 2019

● Improved emotion recognition accuracy for 5% by developing with Pytorch an original deep learning model.

● The model combines EEG signals and eye-movement data, and successfully extracts shared features and individual features. These features provide new insight on the relationship between emotion and brain activity. School of Information, UC Berkeley Berkeley, CA, USA Research Apprentice, AskOski project February, 2019 - June, 2019

● Improved model prediction accuracy of course enrollment for 3% by setting up hyperparameter tuning pipeline and developing with Keras hyperparameter tuning code for RNN.

● Optimized training process by testing how different training and tuning periods affect model performance, and found the period that minimizes computing cost and maximizes prediction accuracy. PROJECTS

Quora Question Classification December, 2018 - January, 2019

● Using Python and Keras, built a machine learning classifier which labels Quora questions as sincere or insincere, and achieved an F1 score of 0.70 on a test set of 50,000 questions.

● Corrected misspelled words and cleaned punctuation with NumPy on a training dataset of 1.3 million questions.

● Implemented modifications (ie. adding attention layer) on multiple deep learning models such as CNN and RNN.

● Performed hyperparameter tuning and ensembling to enhance F1 score by 0.05. B+ Trees February, 2020 - February, 2020

● Implemented in Java complete B+ tree indices, supporting searching, inserting, deleting and bulk loading of keys, and rebalancing after each operation.

World Simulation April, 2019 - May, 2019

● Developed in Java both the UI and the back end of a 2D tile-based world exploration engine, where the users can explore a randomly generated world through keyboard inputs.

● Added features so that users can interact with different objects in the world, and replay their most recent moves.

Contact this candidate