猿急送>

其他其它兼职程序员

ID：162168

wilson

量化工程师

公司信息：
ETrade Inc

工作经验：
3年

兼职日薪：
1000元/8小时

兼职时间：
下班后
周六
周日
可工作日远程

所在区域：
其他
全区

技术能力

• Built models and designed automated algorithms for farm insurance pricing through Python with 5 team members.
• Applied seasonal time series model to predict farm product prices based on past data to determine strike prices.
• Optimized diversified efficient portfolios of a set of farm insurance and simulated efficient frontier using Monte
Carlo process with one million iterations, delivering insights and strategies of investment to clients.
• Created real-time interactive visualization to help clients customize own portfolios and calculate possible return.

项目经验

Analyzed the influence of the number of cars produced every year on Dow Jones U.S. Auto Manufacturers Index using
multiple machine learning methods.
• Collected vehicle production amount from companies' annual reports and DJUSAT index, along with some other
macroeconomic index from Bloomberg; built models by using machine learning techniques like Logistic Regression, Lasso, SVM, Gradient Boosting and compared the results;
• Tested the significance of production amounts, especially the amounts of famous brands.
Investigate the SQ3R study system's effectiveness Jan. 2019-Mar. 2019
• Analyzed the relationship between the time students spent on the SQ3R study system and students' academic performance.
• Cleaned panel data, including dealing with missing data, checking the existence of overlaps and gaps in time frames, examining the identical of constant variables for every ID.
• Analyzed the cleaned data, and reached the conclusion through summarized tables.
Investigate the Performance of Online Learning Models in Security Market July. 2017-Sep. 2017
• Built online learning models with 300 stocks included in CSI300 index and compared the performance of securities portfolio
managed by models with the real CSI index values.
• Collected CSI300 index and its 300 A-share stock prices between Jan. 1 and June 30 from Wind; used online gradient decent
method to build decision models and manage the portfolios; compared the return of the portfolios using OGD method and the corresponding return of real CSI index and obtained the conclusion that OGD performed slightly better than the market.