This repository documents all my data science projects, where I explore different machine learning algorithms for predictive analytics. Built by Debby Ling

• Used RandomForestRegressor to predict the top 11 contestants of Produce 101 Season 2 to achieve accuracy of 82.9%, from 5 weeks of data.

"Produce 101 Season 2 (프로듀스 101 시즌 2) is a 2017 reality boy band survival show on Mnet. It is a large-scale project in which the public "produces" a unit boy band by choosing members from a pool of 101 trainees from 54 entertainment companies as well as the group's concept, debut song, and group name. 11 trainees will be selected to form the unit boy group." Wikipedia

• With Support Vector Machine, predicted with 99% accuracy whether a donor would donate blood in the subsequent month.

• Nature or nurture -- will the education level of parents influence that of their child? Is the education level of a father a greater determinant? Guardian and conventional wisdom affirms this. Using simple linear regression, I predict with 97% accuracy the results of students.

• With Artificial Neural Networks, predicted the state of wells in Tanzania using artificial neural networks and xgboost with a 70% accuracy.

• Nutella's change in recipe was a public relations disaster. Or was it? Although the news was first revealed on Facebook on 2 Nov, most of social media didn't quite react until a Daily Mail article was published on the 6th. While Nutella did take a hit in the positivity of sentiments on Twitter that day, sentiments on Twitter recovered quickly.