Kickstarter Campaign Analysis to Predict Outcome & the Percent of Fundraising Goal to be Raised

Explores 66D feature set from a public Kaggle dataset of Kickstarter campaigns. Investigates whether a statistical model could be fit to accurately predict the outcome and percent of funding goal that would be achieved by a campaign's end.

View the Jupyter Notebook for this project

Modeling techniques used

Predicting campaign outcome (success / failure)

• k-Nearest Neighbors (KNN)
• Decision Tree

Predicting percent of goal achieved by campaign's end

• k-Nearest Neighbors (KNN)
• Stochastic Gradient Descent (SGD)
• Ridge Regression (Tikhonov Regularization)
• Linear Regression
• Lasso Regression