Portfolio
Machine Learning
Customer Churn Detection
I performed comprehensive EDA to identify the portrait of a churned customer, handled missing values, performed feature engineering, and implemented different classification models to predict churned customers, executed model validation to compare the performance of these models.

Anomaly Detection using Kmean++
I implemented PCA to reduce dimensions for better interpretation, and applied Kmean++ to identify the abnormal cluster in temperature sensors data. Compared Elbow and Silhouette scores to choose the optimal number of clusters.

Daily Data Science
Utilizing Natural Language Processing to filter new vocabulary from a Youtube subtitle
I built a program to facilitate my English learning journey by crawling subtitles from Youtube videos, utilizing NLP to do text preprocessing, and developing a graphic user interface to offer solutions to interact with my program.

Identify who is the most productive productivity-Youtubers

After using Python to get data from YouTube API, I start doing EDA analysis in PowerBI to answer my curiosity: Who is the most productive productivity-YouTubers?
Template Source