Name: Yifan Yuan
Email: [email protected]
Github: http://github.com/tsingjyujing/
- Jiangsu University
- 2011.09 to 2015.06
- Course Degree: Bachelor of Automobile Engineering
- Position: Machine Learning Engineer
- Time: 2019.05 until now
- Position: Big-Data Department Manager
- Time: 2017.12 until 2019.03
- Leaded big-data group to build batch computation ETL system by Apache Spark and Apache HBase;
- Designed storage system of data-warehouse by using MongoDB cluster;
- Built data visualization system by cooperating with front-end group.
- Selected technology framework to develop, we use Apache Kafka and Apache Flink to processes hight TPS real-time data (about 2k~20k TPS);
- Leaded team to build streaming computation ETL system.
- Preprocessed data from data-warehouse.
- Accomplished driving behavior classification model by analysis claim data, we used logistic model to predict the risky of vehicle by it's driving behavior features in data-warehouse;
- Built real-time prediction system by using Apache Flink and write a Django application to provide service.
- Database middleware developing (Based on MyCat);
- Built GIS system by Spring Boot + GeoScala
- Position: Data Analyst
- Time: 2015.9 to 2017.12
- Extracted features (location/temperature/CAN data/etc..) from database for analysis;
- Analyzed the relation between the extracted features and maintenance records by logistic regression;
- Predicted the damaging risky of running vehicles.
- Feature engineering, extracted features data from large scale data and do preprocessing;
- Constructed an model (based on K-means and Metric MDS) to analysis clusters in extracted data;
- Tagged the clusters manually and build predict model to mapping new data into clusters;
- Built HTTP API by Django to provide classify service to front-end.
- Feature engineering, extracted the features that may affect fuel consumption;
- Explored the correlation between fuel consumption and driving behavior by using linear regression.