Time Series Prediction Using LSTM
Linfeng Lu · Lund University Publications Student Papers (Lund University) · 2019
Time series prediction is the use of a model to predict future data based on previously observed data.Time series prediction and anomaly detection is important for many businesses in the world today.In this report, we experiment with three di erent models for time series prediction and anomaly detection on data about mobile message tra c volume and Successful Delivery Rate (SDR).The data are provided by Sinch, which is a telecommunications and cloud communications platform as a service company.We compare and analyze the experiment results and nd that the model architecture that includes a LSTM encoder-decoder could improve the model performance on Sinch data, but not in all cases, due to the di erence of the datasets.