# What method should i use for a time series prediction?

**URL:** https://community.konduit.ai/t/what-method-should-i-use-for-a-time-series-prediction/1086
**Category:** DL4J
**Created:** [December 29, 2020, 9:58pm UTC](https://community.konduit.ai/t/what-method-should-i-use-for-a-time-series-prediction/1086 "2020-12-29T21:58:56Z")
**Posts on this page:** 1
**Page:** 1

<div class="post-metadata">

### Author: ![fletchni](https://avatars.discourse-cdn.com/v4/letter/f/5fc32e/32.png) [@fletchni](https://community.konduit.ai/u/fletchni)
#### Post date: [December 29, 2020, 9:58pm UTC](https://community.konduit.ai/t/what-method-should-i-use-for-a-time-series-prediction/1086/1 "2020-12-29T21:58:56Z")

</div>

I’m using an RNN with LSTM layers, and I am trying to make a prediction on a test set of 160 steps. Currently, after training my model, I call the MultilayerNetwork.output method on an INDArray of my test data’s features. When I do this, it gives a long array the size of my test data set. What method should I use if I want to output only the prediction on the last time step? So, instead of what I have now where it gives the prediction for the label on each step,1-160, I want it to make a prediction for the 161st step.