0 and more projects completed
A data scientist digging into the probabilistic model and machine learning. Experienced Data Analyst with a demonstrated history of working in the retail industry.
0 and more projects completed
An unsupervised Word2Vec model exploring the usage similarity between emoji and English words on twitter. In order to acheive better performance, we ensemble the CBOW algorithm and Skip-gram algorithm. Provided a real-time APP on AWS cloud and collected user feedback to server database.
A question is often asked when I was helping the marketing team developing strategy, which is HOW MUCH CAN I VALUE A CUSTOMER. An ensemble probabilistic model(CLTV) could handle this problem very well which not only able to predict how much a customer will buy in any future periods but also the probability he or she will buy in a time range. Also, by comparing with classical machine learning methods, random forest, OLS, and SVM, the CLTV model has the best performance in prediction RMSE.
The motivation of this local COVID19 tracker came from early March 2020 the situation people kept ignoring the threat of the coronavirus. I, as a data scientist, believe I should remind people how COVID19 was kept growing rapidly. Web-scraped the lastest figures from daily press releases of the LA public health department. After that, I created an interactive map showing the density and trend of local communities. By the local community map, I want to provide relevant information and wake people from getting numb with the large numbers.
A question often challenging marketer in planning media expense, HOW MUCH my media invest contributed to our final sales. I provided evaluation of media expense efficiency in terms of ROI and offered media distribution optimization suggestions.
Responsible for data analysis on key commercial investments, i.e. marketing campaigns, media investments. Also implement and augment statistical analyses and modeling techniques with both internal and external data to support business decision making..
Responsible for local market insight, including customer segmentation, RFM analysis, and MMM model.
Provide data research plans and data analysis results to clients, including Microsoft, Suning Group, etc.
Utilizes regression, classification, and clustering to model real-world structured and unstructured data. Explores Deep Learning techniques, in addition to Spark on AWS.
Studied advanced Statistics topics and techniques at one of the best University in Scandinavia.
Built a solid knowledge foundation in Math and Statistics, also gained a second degree in Ecomomics.
Data couldn't help our business, but science does.
Emoji is an universal language using by various people from different language background.However, writing in Emoji is not easy. I asked help from machine learning.
Marketers often curious about how much they can get from a specific customer in the future. An ensemble probabilistic model could help them with it.
In early March 2020, people kept ignoring COVID19 and cannot understand the threat from the coronavirus due to lack of access to relavant and precise information. As a data scientist, I created a local community tracker on map.
Data means nothing, until it meet science.
Los Angeles, CA