Tuesday, 9 January 2018

DeepAR Algorithm can be utilised for model training in Amazon Sagemaker

DeepAR Forecasting Algorithm can now be used for model training in Amazon SageMaker. DeepAR is an algorithm that builds precise forecasts by learning the patterns from time series over various large sets of training data with relevant time-series. The deepAR algorithm studies the similarities on various related items in the dataset to provision more precise forecasts. This process enhances upon common forecasting methods like Autoregressive Integrated Moving Average models or rapid change smoothing which treats each time-series independently. By utilizing this shared information on all related time-series, DeepAR can be implemented to a number of time-series challenges such as forecasting traffic to servers or web pages, predicting future product sales to enhance supply chain management, estimating future electricity consumption at the individual household level. 

No comments:

Post a Comment

AWS’s growth story

AWS’s growth story - numbers don't lie! AWS took 123 months, a little over 10 years, to grow to a $10 billion business. Then took only 2...