Yuxin Wang

• Careful observation: Pay attention to details and be able to quickly catch problems and deficiencies in work and correct them in a timely manner.
  
• Strong sense of responsibility: Always have a high sense of responsibility for assigned tasks, ensuring that work is accurate and completed on time.
  
• Communication skills: Able to communicate effectively whether with colleagues, superiors or clients, ensuring accurate and timely communication of information.
  
• Continuous learning: Always maintain the desire for new knowledge and skills, constantly improve yourself, and adapt to the changing work environment.
  
• Data analysis: Have the ability to deeply understand data, accurately interpret the information behind the data, and use statistical and analytical tools for data processing and mining.
  
• Technical proficiency: Proficient in data analysis tools and programming languages, SAS and SQL, and constantly learning and following the latest data analysis technologies and methods.

Project 1:  Retail Performance Monitoring: Create monitoring reports for a marketing campaign:

  To predict customer response rates:      

• Check and manage the data, to find the overall response rate from 193,728 customers ;
• Create missing value indicators for inputs that have missing values. (By SAS: Array);
• Impute missing value to a new dataset; (By SAS: Proc Stdize);
• Split the imputed dataset into training and test datasets. Use 70% of the data for each dataset role. Stratify on the target variable; (By SAS: Proc Surveyselect);
• Build a logistic regression model with the FAST Stepwise method;
• Achieved the ROC is 60.36%, and Somers' D is 20.7%;
• Finally, to test and make sure there is no over-fitting between training and test datasets;

• Project. 2: A car insurance marketing campaign.
• Check and import the data, to get the total response rate; (By SAS: Proc Freq) from 9,134 customers;
• Break down the customer base into multiple segments and compute KPIs for individual segments；
• Analyze how these response varies by different EmploymentStatus groups, MaritalStatus groups, and SalesChannel groups;
• Identify the most important factors which affect response rate;
• Converted the response into numeric by creating a variable; And check missing rate for numerical variables;
• Build a logistic regression model: Split the data into train(70%) and test(30%) and check average response rate;
• Use stepwise to decide the best model, and achieved the ROC is 67.25%, and Somers' D is 34.5%;
• Finally, to test and make sure there is no over-fitting between training and test datasets;

• Online Marketing
• Inventory Coordination
• Digital Marketing Strategy
• Data Analytics
• Microsoft Office -Word, Excel, PowerPoint
• SAS-SQL

• SAS Global Certification Program: SAS Certified Professional Advanced Programming Using SAS9.4

• Chinese (Mandarin);
• English;