Analyze and Predict the Salary of Machine Learning Workers
Project description: More intellectual workers start to learn or even apply machine learning algorithms to solve their business problems. It’s interesting to find out how value what we learned is and which theme of machine learning is paid best.
Project responsibility: 1. Analyzed and sorted out the data sets of the salary questionnaire for machine learning workers on Kaggle, and conducted data cleaning. 2. Extracted the questionnaire of "the most frequently used algorithm in a work" as the data characteristics. 3. Used the LASSO and RLS regression algorithm to predict the salaries of machine learning workers in different fields. 4. Used PCA algorithm to reduce the data and try to use different data features to get better fitting results. 5. Integrated the analysis results into PPT.
2018-3 - 2018-3
Project description: With more and more serious traffic congestion phenomenon, many cities begun to limit vehicle traffic by limiting the number of vehicle licenses, which has led to a new way of buying a license plate: bidding for a license plate. Due to the huge number of bidders but limited number of license plates, we want to use the knowledge of data analysis to predict the best bid time and the lowest bid price for the bidders.
Project responsibility: 1. Interpreted the mechanism of license plate auction and think about the prediction scheme. 2. Collected the relevant data in the official web site, and found out a large number of data about the license plate auction in ShangHai in recent years. 3. Perform data cleaning with EXCEL and conduct basic analysis of data. 4. Establish ARIMA model for time series with SAS, and predict the price of license plate next month.