Activity
Experience & Education
Licenses & Certifications
Courses
-
Big Data Seminar
BIA810
-
Data Warehousing/Business Intelligent
MIS636
-
Experimental Design II
BIA654
-
Knowledge Discover & Data Mining
CS513
-
Marketing Analytics
BIA672
-
Multivariate Data Analysis
BIA652
-
Process Optimization and Analytics
BIA650
-
Risk Management
BIA670
-
Social Network Analysis
BIA658
-
Statistical Learning & Analytics
BIA656
-
Web Analytics
BIA660
Projects
-
Repeat Restaurant Booking Prediction
• Cleaned and normalized 116,978 booking records made on EZTABLE.com (a restaurant-reservation website) between 2012-2014
• Built models to predict the repeat bookings using four different algorithms, including KNN, ANN, CART and Naive Bayes
• Evaluated the models using different statistical measures, such as accuracy, sensitivity and specificity, etc. -
Using News Article to Predict Stock Price Trend
Constructed a decision support system (DSS) that predicts stock price trends after the release of news
- Labeled each article in a training set of news as "up", "down" or "neutral" according to the movement of the associated stock price
- Trained different text classification models and compared their performances
- Develop a web application to help users automatically predict the stock price trendOther creators -
Web Analysis for MoocList
-
Social Network Analysis for YouTube Videos
• Used R and Gephi to visualize the social network map among users who commented on YouTube car videos ads
• Made targeting recommendations for video advertising campaign by identifying and analyzing the network’s structure
-
Predicting the depression severity by postings
-
• Wrote complex SQL queries using complex joins, grouping, aggregation, nested subqueries, etc. to extract the postings and depression test data for each user from SQL Server database
• Extracted features from text data and labelled the users by depression test scores for modeling using Python
• Trained a naïve Bayes classifier and evaluated the performance of the model on the test data set
• Improved the model by selecting new features and new classifiers, which resulted in 9%…• Wrote complex SQL queries using complex joins, grouping, aggregation, nested subqueries, etc. to extract the postings and depression test data for each user from SQL Server database
• Extracted features from text data and labelled the users by depression test scores for modeling using Python
• Trained a naïve Bayes classifier and evaluated the performance of the model on the test data set
• Improved the model by selecting new features and new classifiers, which resulted in 9% increase in AUC
Languages
-
Chinese (Mandarin)
Native or bilingual proficiency
-
English
Professional working proficiency
People also viewed
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Fei Feng in United States
28 others named Fei Feng in United States are on LinkedIn
See others named Fei Feng