Icon

Multivariate_​Time_​Series_​RNN_​Keras_​Training

Multivariate Time Series Analysis with an RNN - Training

This is a simple example workflow for multivariant time series analysis using an LSTM based recurrent neural network and implemented via the KNIME Deep Learning - Keras Integration.

It is based on the bike demand predition dataset from Kaggle and trains a model to predict the demand in the next hour based on the demand and the other features in the last 10 hours.




Define network Preprocessing Train and apply the network Extract predictions and evaluate performance Input shape: 10,1310 steps13 dimensions per step100 unitsReLU with 1 unitEpochs: 75Loss: MSE50 % validation50% testing80 % training20% test+validationDay of monthDay of weekMonth HourRead london_merged.csvfrom kaggleSave model Keras Input Layer Keras LSTM Layer Keras Dense Layer Keras NetworkLearner Partitioning Keras NetworkExecutor Numeric Scorer Normalizer Denormalizer Column Appender Line Plot (Plotly) Partitioning Normalizer (Apply) Extract Date Fields RestructureTraining Set Preprocessing Testand Validation Set Denomrmalize Predictionsand Round to Int CSV Reader Keras NetworkWriter Model Writer Define network Preprocessing Train and apply the network Extract predictions and evaluate performance Input shape: 10,1310 steps13 dimensions per step100 unitsReLU with 1 unitEpochs: 75Loss: MSE50 % validation50% testing80 % training20% test+validationDay of monthDay of weekMonth HourRead london_merged.csvfrom kaggleSave model Keras Input Layer Keras LSTM Layer Keras Dense Layer Keras NetworkLearner Partitioning Keras NetworkExecutor Numeric Scorer Normalizer Denormalizer Column Appender Line Plot (Plotly) Partitioning Normalizer (Apply) Extract Date Fields RestructureTraining Set Preprocessing Testand Validation Set Denomrmalize Predictionsand Round to Int CSV Reader Keras NetworkWriter Model Writer

Nodes

Extensions

Links