Chenwei Wu Email & Phone Number

Ann Arbor, Mi, Us

New Haven, Connecticut, United States

Ann Arbor, Michigan, United States

PhD in EECS, Coadvised by Dr. Liyue Shen and Dr. Zhongming Liu,Multimodal Machine Learning, AI for medicine, AI for neuroscience, Meta-Reinforcement Learning

Collaboration with Dr. Leo Anthony Celi at the MIT Laboratory of Computational Physiology, Medical AI Fairness and Privacy, Medical Large Language Models

• Develop real time gaze detection (Gaze360) and facial recognition algorithms via Pytorch for Harvard Makerspace and integrate the tracking system into the Multimodal learning analytics cloud data pipeline. • Collect data of student behavior in Makerspace lab sessions by leveraging the Multimodal learning analytics pipeline, and perform analysis to understand social learning aspects like collaboration and student attention.

• Develop viewmaker networks, which are generative models with stochastic boundaries for data augmentations, via Pytorch Lightning, to adversarially auto-learn and generate augmentations on 12-lead electrocardiogram (ECG)sensor data for contrastive learning tasks, so as to reduce the rigorous trial and error by human experts.• Investigate and compare the performance of viewmaker networks to those of other previous contrastive methods, in particular whether viewmaker networks learned views that are medically sensible, and whether they are more robust to corruptions commonly observed in ECG data collection settings.

Rochester, New York, United States

• Performed sentiment and topic analysis on text behavioral data in a functional neuroimaging study to better understand the relationship between brain functional connectivity and social anhedonia. • Applied machine vision techniques including semi-supervised CNNs, optical flow and open face, on video datasets to analyze group differences in nonverbal synchrony during social interactions among people with schizophrenia and controls. • Studied patterns of combinations of facial actions to evaluate the effect of oxytocin on patients' social abilities.

Rochester, New York, United States

• Utilized unsupervised machine learning algorithms to cluster nursing homes based on the percentage of residents with dementia, depression, and serious mental illness, and detected previously unknown patterns of resident case-mix and staffing in nursing homes.Predicted deficiency scores of nursing homes on a longitudinal basis with supervised learning algorithms.• Developed text association and pattern mining algorithms to classify cancer therapies. Combined synthetic minority oversampling Technique (SMOTE) with supervised learning techniques to deal with unbalanced caregiver datasets and help clinicians identifypotential mental and physical health risk factors for caregivers of the elder people. Implemented Local Interpretable Model-Agnostic Explanations (LIME) to give a non-black-box explanation for ML results in a clinical setting.

Rochester, New York, United States

• Processed data from anagram tests among suicidal adolescents to examine level of anxiety and cognitive flexibility.• Applied text processing techniques via Python to extract patient information from electronic medical record notes of adolescentsadmitted for suicidal crises across a variety of care units at University of Rochester Medical Center.• Developed scripts in Python to consolidate inclusion/exclusion criteria data across multiple reviewers for a set of 1700 empiricalarticles examining self-injury and suicidal thoughts and behaviors among gender minority adolescents.

- Assist teaching in an upper-level data science course, Data Mining - Grade homework, exams, and projects (implement Apriori and FP-growth tree algorithm, Advanced Regression Techniques, Classification) and hold weekly office hours

• Develop the optical components of a digital eye tracker and implement experiments to test the influence of eye movement on vision.• Perform data mining on eye movement data and develop a software for data acquisition and processing.

New York City Metropolitan Area

Served as developer & business analyst for Investment Banking Division, building Airflow automated data ETL pipelines and constructing a centralized Azure cloud data platform for bonds and credit default swaps.

• Form a Data Science Colloquia Series, which will involve inviting renowned speakers from varying fields of data science to disseminate modern topics and methods.• Provide Data Science major students with internship and research opportunities by cooperating with faculty and utilizing alumni resources in the industry.

New York City

• Served as project manager for the PoC of a firm-wide chat platform that leverages NLP to assist sales & trading team to a competitive edge. • Collected big data streams and performed deep learning time series predictions on stock trends.

Shanghai, China

• Organized surveys on Beijing’s business and investing environment and performed data preprocessing for the survey to provide theGovernment with a better view of the Beijing market and help policy making.

• Assisted teaching for an upper-level economics course, Economics of Globalization • Held office hour every week answering inquiries and explaining economics concepts to students

Computing with The Community Program• Provided STEM education via hardware and video game creation among urban youth seeking a high school equivalency diploma.• Developed an Android App local data to reduce risks associated with alcohol use in social settings.

Boston, Ma

• Analyzed certification and accounting data via R and Excel VBA to identify opportunities and make strategies for the recruitment, engagement, and retention of minorities, women, LGBTQ, physically disadvantaged and veteran vendors. Actively attended off-line events and make calls to promote the government help programs and re-establish connections with disadvantaged vendors. • Utilized BI reporting tool (Sugar) for data gathering, analysis and visualization; Researched and utilized queries via MSAccess to quantify relevant data within the Commonwealth’s Data Warehouse.

• Developed a new car seat that automatically adjusts positions and alarms under potentially dangerous driving conditions.• Cleaned data collected from sensors and mined patterns of seating preference, body temperature and movement from the data tocreate personalized user experience.

•Coded an AI for a robotic arm that could play table tennis and won first prize in the 4th Omron Sysmac Design Competition.

China

• Learned ICBC’s personal and corporate banking products and services such as “Link Express” series security products and WeChatBanking services.• Responded to inquiries, assessed customer needs, and provided suggestions on bank services and products accordingly.

Doctor Of Philosophy - Phd, Eecs, 4.00/4.00

Master Of Science - Ms, Data Science, 3.94/4.00

B.S. In Data Science, B.A. In Financial Economics, Minor In Japanese, Gpa 3.92/4.0, Phi Beta Kappa