Network Traffic Prediction using Quantile Regression with linear, Tree, and Deep Learning Models

摘要: Machine Learning research has progressed tremendously in recent years. Major fields that machine learning pushed its frontier were prediction and data modeling. In this work we evaluate the applicability of a handpicked models on predicting inter-day aggregate network traffic. We chose best with multi-variate feature space. They represent linear, decision trees, neural models. Over years, traffic resorted to point values. This approach is not descriptive enough naively gives shallow conclusion about data. propose using quantile loss function predicts boundaries or intervals. Our results show linear fared well compared their simplicity while Long Short-Term Memory Neural Networks gave across all experiments.