This document discusses a statistical approach for classifying and identifying DDoS attacks using the UCLA dataset. It proposes extracting features from network traffic such as packet count, average packet size, time interval variance, and packet size variance. A packet classification algorithm first classifies packets as normal or attacks. For uncertain cases, a K-NN classifier is used. Then the types of DDoS attacks, including flooding and scanning attacks, are identified based on the feature values. The proposed approach is evaluated using the UCLA dataset and shows mathematical calculations for feature extraction. In conclusion, the statistical approach and packet classification algorithm are effective for classifying common DDoS flooding and scanning attacks.