Ya Liu
Analytical thinker with the ability to mind hidden gems located within large rich data sets. Capable of turning dry analysis into an exciting story that drives business critical decisions.
View My LinkedIn Profile
Ya’s Portfolio
R project
Titanic: Machine Learning from Disaster
In this project, I trained a model using the Titanic dataset from Kaggle for predicting whether a passenger may survive the sinking of Titanic. The project is carried out in three parts: data exploration, feature engineering and model training. Performance of final model is 81%.
Online Food Trend Analysis
Early detection of emerging food trends can translate into great business opportunities. Today, a lot of food-related discussions occur on social media platforms such as Twitter and Facebook. Thus, such social media content presents a potentially valuable and real-time source of intelligence that can be leveraged by retailers to better serve its customers. The purpose of this project is to explore this possibility using techniques discussed in the social media analytics course to help retailers see the rise and fall of certain categories of food before competitors do.
The project is based on 4 million Facebook posts from 2011 to 2015. Here I will validate my method with the case of Cauliflower Rice whose demand has been growing steadily from 2011 to 2015.
Analytics Design to Recommend Flavors to Launch
In this project, I am conducting analytics to provide recommendations to the decision of which flavors to launch next. Assume for this analysis that the private label recently launched six flavors–Blueberry, Honey, Peach, Plain, Strawberry, and Vanilla. I focused on providing analytics to help determine what the next flavors to launch should be. The main methology I use is TURF analysis.
Python project
Decision Support System
This decision support system is designed to help bank managers predict risk performance of a credit card aplicant. In this project, I trained the predictive model with dataset first and then visualized the system with Streamlit. Bank managers not only can get decision support but also explaination from this system.
Sentiment Analysis on Twitter Airline Dataset
In this project, I trained a sentiment analysis model with the US Twitter Airline Dataset which contains 1700 Tweets on complaint about Airlines and 1700 Tweets not complaining about Airlines. We can tell the sentiment of tweets with model we developed. The precision of our model on validation set is 0.58.
HTML project
Website Prototypes Development
This prototype is part of web analyst project I did for a local heathcare company. It took me two weeks’ rest time to develop this prototype. The prototype provides support for following usability test. Also, it visualized part of recommendations of our project.