Chenyu Liu work email
- Valid
Chenyu Liu personal email
Success in this field requires full life cycle experience in planning and assessment, requirements definition, features and functionality design and development, coding, testing, QA, implementation, product and infrastructure enhancements, and maintenance. I am capable of coordinating the full lifecycle development of next-generation applications, from conception to design, coding, testing, documentation, and implementation. Backed by my Master of Science in Robotics, I pride myself as a Skilled Machine Learning Engineering equipped with years of experience in machine learning, reinforcement learning, motion planning, causal inference, and software development. I have proven myself to be a highly motivated machine learning engineer with the ability to remain focused in high-stress environments. Through my academic experience, I am well versed in implementing machine learning algorithms, improving data efficiency, performing extensive research, and working independently and as a team to perform specified tasks– to name a few; in driving absolute technical deliverables. I possess the robust capacity in analyzing the current process/product, interfacing directly with clients to draft specifications, define requirements, determining gaps, and implementing solutions to maximize efficiency and generate revenue growth. I can define new product releases utilizing deep insights and the latest tech to provide meteoric improvements while communicating to teams and all levels of stakeholders using exceptional communication skillsNotable career highlights include:• Implement machine learning models for item classification and action recognition, evaluate model performance and monitoring the training process as well as adapt existing models to better handle problems specific to our domain.• Demonstrate expertise in cross-platform proficiency, fluency in multiple programming languages, and advanced knowledge of diverse applications, databases, infrastructure architectures, methodologies, full-stack development techniques, and best practices• Knowledgeable in software engineering best practices and dedicated to driving innovation with the ability to follow industry and technological trends and facilitating early adoption of innovations.• Adept in the utilization of programming languages - Python and C++ -software and sophisticated business tools targeted towards business advancement
-
Applied Macine Learning Scientist启元世界 Jun 2022 - Apr 2024中国 广东省 深圳 -
Applied Machine Learning ScientistTeladoc Health Jun 2021 - Jun 2022美国 加利福尼亚州 山景城 -
Project- Reinforcement Learning In Motion PlanningDuke University Apr 2019 - Dec 2020Durham, North Carolina, United States• Added to the body of knowledge by collaborating with researchers to publish a conference paper that focuses on applications of causal inference on reinforcement learning while being the main author.• Provided analytical solutions by developing numerical simulations in Open AI Gym/Unit 3D Engine.• Achieved efficiency by implementing and validating proposed algorithms in built simulations, setting up baseline algorithms in the same environment for comparison, not limited to imitation learning, hierarchical learning, generative models, clustering and PPO (TRPO).• Designed algorithms that improved data efficiency and solved problems of reinforcement learning in motion planning problems. -
Masters StudentDuke University Jan 2019 - Dec 2020
-
Project-Path Classification By Unsupervised LearningDuke University Apr 2020 - Jun 2020Durham, North Carolina, United States• Given a set of trajectories generated by an agent in a motion planning problem (vehicle movement or robot manipulation), unsupervised cluster the trajectories into several patterns of behaviors.• Selected and implemented state-of-art unsupervised learning algorithms alongside deep-clustering, generative models (VAE, Info-gail) which achieved 96% in trajectories classification problems.• Utilized TensorFlow to execute algorithms and deep-learning framework.
-
Project-Build 3D Motion Planning Mini-Game As Reinforcement Learning Simulation EnvironmentDuke University Jan 2020 - Mar 2020• Used Unity 3D Engine to build a 3D robot navigation mini game in which an external neural network can be trained and manipulate the robot in the game to finish motion planning problems in a dynamic environment.• The game was able to collect and export all the state-action data of the in-game robot during interaction which made it a suitable environment for transfer learning algorithms.
-
Project-Transfer Learning In Contextual Multi-Arm Bandits With Biased Observational DataDuke University Oct 2019 - Dec 2019• Solved the problem where the biased observational data is given in transfer learning of contextual multi-arm bandits problem. • Designed and implemented causal-bounded UCB-1 algorithm (with Python) that avoids biased transfer in such problem settings and greatly improved the data efficiency.
-
Project-Transfer Learning Under Different Action SpacesDuke University May 2019 - Sep 2019• Designed algorithms which solved transfer learning in motion planning problems where the demonstrator and learner agent had different actuator.• Implement designed algorithm with C++ in discretized simulation, based on discrete space RL algorithms, such as Q-learning and SARSA.
-
Research AssistantHumans And Autonomy Lab, Duke University May 2019 - Aug 2019Durham, North Carolina, United StatesProjects: Implemented innovative blueprints that facilitated the completion of the Drone Project Design; developed drone tracking system and testing environment; achieved optimum performance by building control software and piloting software for quadcopter drone.• Worked with team to gather information and data relevant to the successful completion of the projects; also completed projects in a timely and efficient manner.• Provided extensive assistance with laboratory analysis, quality control, data management, problem identification and problem solving.• Prepared and presented reports on tests performed, results obtained and other activities to enhance project completion.
-
Research AssistantInstitute Of Semiconductors, Chinese Academy Of Sciences May 2018 - Aug 2018Beijing, ChinaProjects: Programmed in C++ to simulate the synthesis process of perovskite solar cell, which eliminated unnecessary experiments and chemical combinations, reducing the unnecessary experiments conducted by 10%.• Worked with team to gather and process data related to specified projects and research work; also utilized appropriate data analysis method to derive insight from gathered information.• Efficiently utilized database programs to maintain and update records pertaining to project findings and population; verify data to ensure accuracy of research finding and eliminate errors which may influence research result.• Provided extensive research support by preparing experiments and analyzing data; also assisted students with their research work by supervising them.
Chenyu Liu Skills
Chenyu Liu Education Details
-
3.85
-
Material Science & Engineering -
Earth Science
Frequently Asked Questions about Chenyu Liu
What is Chenyu Liu's role at the current company?
Chenyu Liu's current role is Applied Machine learning Scientist.
What is Chenyu Liu's email address?
Chenyu Liu's email address is cl****@****doc.com
What schools did Chenyu Liu attend?
Chenyu Liu attended Duke University, The Ohio State University, Sun Yat-Sen University.
What skills is Chenyu Liu known for?
Chenyu Liu has skills like Deep Reinforcement Learning, Gaming Ai, Algorithms, Causal Inference, Numerical Simulation, Reporting And Analysis, Reinforcement Learning, Matlab, Troubleshooting, Deep Learning, Recommender Systems, Artificial Intelligence.
Not the Chenyu Liu you were looking for?
-
Chenyu LIU
B2B Channel Management, Sales Strategy Planning & Operation @Alibaba | Msc. @Hku | B.S. In Mgt @HitHangzhou -
CHENYU LIU
King'S College London Ma Master Of Arts In Teaching English To Speakers Of Other Languages (Tesol) Ielts | Alevel | Eap | IgcseChengdu -
1flextronics.com -
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial