Shenghan ZHANG

 Already done: Data Structure Learning, developed python and java ability to advanced level, better data visualization skills and data analysis ability, trained multiple Psychometric Tests.  Learning: coursera NoSQL Certificate, Machine Learning Certificate.

CODALAB PROJECT “Contextual Emotion Detection in Text” . The project aimed to build an emotion recognition system that can recognize the emotional changes of chatters in online chat conversations.  Machine Learning & Data Analysis: Used 10 deep learning model Variants through Tensorflow and Keras based on LSTM, GRU, CNN for supervised learning. After model training, it finally reached precision 91.1% and recall rate 89.5% in the recognition of dialogue emotions.

Team Project “Integrating Product Data from Web” (School-Enterprise Cooperation Project) The project aimed to compare multiple commonly used products about its pricing on different websites.  Model Training: Supervised Leaning Models were performed on the obtained data. Through comparison with 2,000 manually labeled products, finally achieved the precision 84.3% and the recall 76.5%.  Data Visualization: Used Tableau to visualize data, to provide data support for the final project implementation.

Marketing Analysis: Analyzed the operation results of corporate hedging, and provided the research department with an analysis report on the willingness of the company to participate in the project. Helped the research department improve the performance of project by about 14%.

 Data Analysis: based on data from Baidu Map (relevant to Google Map) analyzed customers preference in B2C industries.  Sales Performance: through providing reports, cold calls and visits, finally reached 40+ effective customers every day, and signed about 50+ customers every month, achieved the top 15% of the department.