Jui-Hung (Ray)Lu
Location: Corvallis, OR, *****, U.S.A Phone: 702-***-**** Email: *****@***********.*** LisnkedIn: https://www.linkedin.com/in/jui-hung-lu-92001613b/ CAREER OBJECTIVE
Enthusiastic and highly motivated software engineering graduate student with proven problem-solving and algorithms understanding capability. Experience in Agile management, software application development with CI/CD system, data analysis, and learning algorithms implement.
EDUCATION
Oregon State University (OSU) - Corvallis, OR Apr 2018 - Dec 2019(expected) Master of Engineer, Computer Science, GPA: 3.77
Courses: Algorithms, Database System, Machine Learning, Deep Learning, Reinforcement Learning National Central University (NCU) - Taoyuan City, Taiwan Sep 2012 - Jun 2016 Bachelor of Science, Communication Engineering
Courses: Algorithms, Data Structure, Computer Communication Networks WORK EXPERIENCE
Field Application Engineer Feb 2017 - Jun 2017
GCOM TECHNOLOGIES CO., New Taipei City, Taiwan
• Be responsible for engineering a system in production, helping with product testing and managing Research Assistant Feb 2015 - Jun 2015
National Central University, Taoyuan City, Taiwan
• Provided three approaches using Java and Android application to evaluate video streaming transmission delay
• Implemented real time video transmission in no Internet environment using Wifi- AP mode or Wifi-Direct mode PROJECT EXPERIENCE
A Cloud and Flask based Real Estate Predict system AWS, Docker, Jenkins, Python, Restful, Flask Jul 2019 - Sep 2019
• Developed a machine learning application with 70% accuracy and automatic web crawler for auto data updating and predicting using SoupBeautiful in Python
• Designed backend server and deployed on AWS EC2 using RESTful API in Flask framework
• Incorporated application and server to a CI/CD pipeline system using Docker, Jenkins which automated steps in software delivery process and reduced 70% periodically updating work Image Feature Detecting Improvement Research PyTorch, TensorFlow, CNN, SIFT Jan 2019 - Mar 2019 Handcrafter Descriptors vs Learned Descriptors
• Improved accuracy to 97% by integrating Handcrafter and Learned Descriptors, which accuracy is 23 % higher than Learned Descriptors Model and 32% higher than Handcrafter Descriptors Model
• Developed CNN models and SIFT models as Learned and Handcrafter Descriptors using Torch in Python Cart-Pole Game DQN, Q-Learning, SARSA Jun 2019
• Implemented Deep Q Network and achieved 50% faster convergence than Q-Learning and SARSA Signal Character Recognition AdaBoost, Random Forest, KNN, Perceptron Dec 2018
• Implemented regression, perceptron, KNN, random forest, and AdaBoost without applying any present model and achieved over 96% accuracy with AdaBoost using Python Restaurant Search Engine and Recommendation System Agile, MySQL, PHP, Html, CSS Sep 2018 - Dec 2018
• Developed product and achieved 90% customer satisfaction by gathering user’s feedback in Agile
• Designed a backend search engine web and database system using HTML, CSS, JavaScript, and MySQL
• Created user certification and provide customized searching result based on user’s record using PHP SKILL
• Programming Language: Python, C, Java, Bash, Haskell, HTML, CSS, PHP, MySQL
• Framework and Libraries: Linux, PyTorch, Restful API, OpenCV, Flask, Android, Scikit-learn, OOD
• Tools: Git, AWS, Jenkins, Eclipse, CI/CD, Docker, DevOps, Agile, TCP/IP, Computer Network Model