PhD, New Jersey Institute of Technology
Dr. Zhitao Wang obtained his Ph.D. degree in Applied Physics at New Jersey Institute of Technology (Feb. 2011 - Aug. 2017). He worked with Dr. Dale Gary and Dr. Bin Chen in the Center for Solar-Terrestrial Research. His doctoral thesis focused on the solar radio observation with state-of-the-art technique of dynamic imaging spectroscopy. During his Ph.D. studies, he has published 3 papers as first author in leading journals, including the Astrophysical Journal and IEEE journal of Photovoltaics.
Zhitao is very experienced in data analysis, using python as one of his primary programming language. His reseach involved analyzing solar radio data, which has a typical raw data rate of about 400 GB per hour for the VLA radio instrument. Nowadays, modern astrophysicists are facing colossal amount of data set per day, and therefore data mining and pattern discovery become an important tool in this field. In particular, one type of radio signatures, the so-called solar fiber bursts, is characterized by its quasi-periodic fine structures frequently appearing for over 1000 times in just about a minute. Using an automatic feature tracing algorithm, Zhitao has successfully detected and traced many of these signatures, from which important information about coronal physics can be obtained. The importance of this work has recently been recoginized by the Astrophysical Journal.
In addition, his interest is not only restricted to the research but also extended towards developing several interesting software applications (see each project in project showcase). During his Ph.D. studies, Zhitao has already taken several critical CS courses, either from the university or online, e.g., data structure and algorithm (CS 610), Java programming (CS 602), machine learning, data base, web development, etc. In one of the graduate courses (Java programming) offered by NJIT in Fall 2016, he ranked 1 out of 35 CS graduate students. Overall, he has shown a good foundation in CS, and demonstrated the potentials of becoming a software developer.
Just after his graduation, Zhitao is now actively seeking for opportunities in software development, preferably in the career track of backend developer, data engineer or data scientist. Please feel free to contact him to see if he is the right fit for your job position.
- Courses/Skills: Java Programming (Class Rank: 1 of 35), Data Structure and Algorithm, Data Analysis, Database, Web Development, OOD, Cloud Computing, Mobile Development, Big Data, Machine Learning
- Tools: Apache Tomcat, MAMP, Eclipse, JUnit, JMeter, Android Studio, MySQL, MongoDB, Amazon EC2, Git/Github, Gradle, XCode, RESTful, ElasticSearch, Google Cloud, BigTable, BigQuery, Dataflow, AdMob, Firebase
Overview: Developed a dynamic web application for users to search events by geolocations, and improved personalized experience by adding event recommendation based on user's preference.
Back End Integration
- Created Java servlets with different RESTful APIs to handle HTTP requests and responses of event search, favorite, and recommendation.
- Built relational (MySQL) and NoSQL databases (MongoDB) to filter event data from external TicketMaster API.
- Used factory pattern to switch between MySQL and MongoDB (more scalable) databases depending on the use case.
- Designed a content-based event recommendatation algorithm based on user's search history and favorite events.
Front End Integration
- Connected the frontend to the backend APIs, and deployed server side to Amazon EC2 to handle 150 QPS tested by Apache JMeter.
Data Log Collection
- Used ElasticSearch to collect logs from remote environment, such as crash report, usages, abnormal behaviors.
- Used Logstash to build a dynamic, real-time pipline to filter and extract information from the logs.
- Visualized logs with Kibana to analyze and visualize the distribution of users' geolocation (View Map).
2. iOS and Go Application Development: Geo-Index Based Social Network
(Demo Movie, iOS Github, Go Github)
Overview: Developed a social network platform for iOS users to post and search messages based on the geolocation.
Back End Integration
- Built a reliable web service in Go to handle posts and deployed to Google Cloud for better scalling.
- Utilized ElasticSearch to provide geolocation-based search functions so that users can promptly query nearby posts/messages within a radial distance (e.g., 200 km).
- Used Google Dataflow to implement a daily dump of posts to BigQuery table for offline analysis.
- Aggregate the data at the user level and post level to improve the keyword-based spam detection with BigQuery.
- Used a realistic dataset of SMS text message with ~5,000 samples to train the spam classifier based on TF-IDF corpus, and obtained an average of 97% precision/recall (Jupyter Notebook).
Center for Solar-Terrestrial Research (CSTR)2013.02-2017.08
- Analyzed solar radio data with a raw data rate of 400 GB/hour under supervision of Dr. Dale Gary (director of EOVSA)
- Implemented a feature-tracing algorithm to track over 1000 solar radio burst signatures, and obtained the mean properties among different groups of burst signatures based on their statistics.
- Applied a gradient descent algorithm to minimize the cost function between the observed data and model, and derived critical information about the solar corona from the best-fit model parameters.
- Helped professors to implement parallel computing in EOVSA software using multiprocessing package in python, and significantly reduced the computation time by 80% on the solar radio data.
- Developed a database application for EOVSA to serve the solar physics community by allowing users to search, and download observing data from the EOVSA website
New Jersey Institute of Technology (NJIT)
Ph.D. in Applied Physics (GPA: 3.70/4.0, 3 papers as first author in leading journals)2011.02-2017.08
M.S. in Applied Physics (GPA: 3.88/4.0, master fellowship)2009.09-2011.02
Guangdong University of Technology
B.S. in Optical Information Science and Technology (GPA: 87/100, ranked top 3%)2005.09-2009.07
Publication and Presentation
- Study of Coronal Active Region and Solar Bursts with Radio Dynamic Imaging Spectroscopy
- Wang, Z., Chen, B., & Gary, D. E., "Dynamic Spectral Imaging of Decimetric Fiber Bursts in an Eruptive Solar Flare," The Astrophysical Journal, 2017. arXiv DOI
- Wang, Z., Gary, D. E., Fleishman, G. D., & White, S. M., "Coronal Magnetography of a Simulated Solar Active Region from Microwave Imaging Spectropolarimetry," The Astrophysical Journal, 2015. arXiv DOI
- Wang, Z., Cheng, Z., Delahoy, A. E., & Chin, K. K., "A Study of Light-Sensitive Ideality Factor and Voltage-Dependent Carrier Collection of CdTe Solar Cells in Forward Bias," IEEE Journal of Photovoltaic, 2013. DOI
Selected Conference Paper and Poster
- Wang, Z., Chen, B., & Gary, D. E., "Dynamic Spectral Imaging of Decimetric Fiber Bursts in an Eruptive Solar Flare," Contributed Talk, 14th RHESSI Workshop, Newark, 2015.
- Wang, Z., Chen, B., & Gary, D. E., "Tracing Solar Fiber Bursts Spatially and Spectrally with Microwave Imaging Spectroscopy," Conference Poster, Joint American Astronomical Society/American Geophysical Union Triennial Earth-Sun Summit, Indianapolis, 2015. ADS
- Wang, Z., Gary, D. E., & White, S. M., "Coronal Magnetography of a Simulated Solar Active Region from Microwave Imaging Spectropolarimetry," Conference Poster, 224th American Astronomical Society Solar Physics Divison Meeting, Boston, 2014. ADS
- Wang, Z., Cheng, Z., Delahoy, A. E., & Chin, K. K., "A New Solar Cell Modeling for CdTe Solar Cell," Conference Paper, 27th European Photovoltaic Solar Energy Conference and Exhibition proceeding, Frankfurt, Germany, 2013. EU PVSEC
Back to Top