An e-learning company runs a website where the users may sign up for free but must pay to access pro features. The company wants to develop an automated system that can forecast if a new user will upgrade to a pro subscription within a year. The company has put together a labeled collection of dataset from two million users. The training dataset contains 2,000 positive samples (users who paid within a year) and 1,998,000 negative samples (users who never upgraded to pro). Each data sample contains 100 attributes about the user such as their age, device, profession, etc. The analytics team at the company developed a random forest model on this dataset, which converged to above 99 percent accuracy on the training set. However, the prediction accuracy on a test dataset was below par. Which of the following solutions would you recommend to the analytics team for addressing this use case? (Select two)