Train multiple anomaly detetction models for serveral multivariate time series

General Discussion
,
#1

Hello ,
I have a dataframe that contains about a thousand multivariate time series I want to detect the anomalies on these time series with autoencoders Is there a way to tune the hyperparameters for each model with tensorflow? If I train all the models with the same hyperparameters can I get good results?
Thanks in advance

#3

Hi @Bacem_STITI

Welcome to the TensorFlow Forum!

Could you please refer to this matching model of Keras - Timeseries anomaly detection using an Autoencoder as per your requirement which might be helpful to you in this issue. Thank you.

1 Like