Predicting the popularity of contents in social networks is quite important for several applications such as viral marketing, news propagation and personalization. In this work, we developed an statistical learning approach to predict the popularity of tweets in the twitter social network. We extracted several user-based, tweet-based and network-based features from each tweet and adopted several classifiers to predict the popularity of tweets. We model this problem with a binary classification problem where popular tweets are considered as the positive and non-popular tweets are considered as the negative class. Popularity is defined by a threshold which indicates how many time a tweet is retweeted. We defined several popularity thresholds and examined the performance of different classifiers based on different threshold values. Our experimental results show that there is no global best classifier for the problem of popularity prediction in twitter but depending on the dataset, popularity threshold and our interest, we can adopt an optimal classifier with a proper set of features for this task.