Ranking

Compare separate action predictors, shared neural networks, and stacked models for feed ranking. Follow the same like, comment, and retweet targets through each design, and assess what computation and information the models share.

Use the training data to predict the probability of a like, a comment, and a retweet for each viewer-tweet pair. These are separate binary classification targets. Compare their predictive quality with validation measures such as classification error or AUC, while keeping the ranking objective in view.

Modeling options

Some questions to consider are: Which model will work best for this task? How should you set up these models? Should you set up completely different models for each task? Let’s go over the modeling options and try to answer these questions. We will also discuss the pros and cons of every approach.

Logistic regression

Initially, a simple model that makes sense to train is a logistic regression model with regularization, to predict engagement using the dense features that we discussed in the feature engineering lesson.