Xinli JiaXinli JiaXinli Jia About Me ProfileEducationSkillsProjectsExperienceResearch Education Skills Projects Experience Research

Name

Xinli Jia

Email

[email protected]

Tel

(1)347-331-9714

Addr

548 Riverside Dr, New York, NY 10027

Education

Columbia University

MS in Electrical Engineering
Expected Dec 2017

Shanghai Jiao Tong University

BS in Electrical and Computer Engineering
Aug 2016

Skills

Programming Language

Proficient:
Python • C++ • Matlab
Familiar:
C • Java • R

Operating System

Linux • MacOS

Web

Django • Ruby on Rails •HTML

Database

AWS EC2 • DynamoDB • Kinesis MySQL • S3

Projects

IN-CAR MUSIC RECOMMENDER

Feb 2017 - Current | Columbia University
• Built a model on AWS machine learning to recommend appropriate music based on in-car environment and other factors.
• Used Kinesis stream to update the model in real time.
• Built a web APP using Django to demo the results.

PREDICTING STOCK MOVEMENT

Sep 2016 - Dec 2016 | Columbia University
• Built a model to predict price trends of particular stocks based on historical movement using Spark Python.
• Used R to clean the data and run the model on databricks.

COORDINATE SYSTEM GRAPHER LANGUAGE

Sep 2016 - Dec 2016 | Columbia University
• Implemented a programming language in Ocaml to plot graph in coordinate system.
• Created a compiler by Ocaml to translate source program to LLVM.

Experience

INTERN - LOGIC SOLUTIONS

Sep 2015-Dec 2015 | Shanghai, China
• Contributed to designing an Apple Watch APP, ActiveFitConference, to bring the athletic lifestyle into the conference rooms.
• Conducted a study and training on APP development on OS X.
• Conducted a review on related products and a market analysis, and
completed a proposal for APP functions and architectures.

Research

CROWDSOURCED INDOOR LOCALIZATION BASED ON WIFI SIGNALS

Jan 2015-June 2016 | Shanghai, China
• Proposed and implemented a crowdsources peer-assisted indoor positioning system based on WiFi signals.
• Adopted a core algorithm of FeetSLAM techniques (Simultaneous Localization and Mapping for pedestrians), and contributed to the algorithm implementation in Java and Matlab.
• Achieved an average positioning accuracy of within 2m performance in different environments.

A NOVEL FACE RECOGNITION ALGORITHM

Mar 2014-Oct 2014 |Shanghai, China
• Developed a face recognition algorithm by combining principle component analysis method(PCA) and Local Binary Pattern(LBP) method in Matlab.
• Demonstrated a recognition accuracy higher than 96%, comparable to many state-of-the-art methods.
• Filed a Chinese patent on this work: ‘X. Jia, B. Li, L. Tian, X. Hao and T. Sun, Patent No. 2014106082274, Oct. 2014.’