This project is to build a model to predict whether a user will download an app after clicking a mobile app advertisement, based on a vast collection of users’ clicking record on app advertisement as our original training data including attributes such as ip, os, device, channel etc. After thorough analysis and preprocessing of the dataset, we selected 6 features including ip, app, device, os, channel, and click_time for training part and set 2 features “is_attributed” and “attributed_time” as output labels for both training and testing parts. We firstly tested Decision Tree (ID3 & C4.5) and obtained a convincing accuracy of 90%. Next, we tested several algorithms in WEKA, including K-Nearest Neighbor, Ada Boosting, Logit Boosting, Classification via Regression, and Random Forest. And the results turned out that Classification via Regression and Random Forest led to better results for this prediction task compared with other algorithms. Finally, we combined MultiLayer Perceptron Classifier with Regressor, the accuracy of which came to nearly 95%. To summarize, we could safely say that the result of combination of MLP classifier and regressor is better than either of them and other used algorithms, which is able to come up to the original goal in our task overall.