univariate single-step style lstm
Tutorial on Univariate Single-Step Style LSTM in Time Series Forecasting
The last layer of the model, the output layer, will predict only one numeric value. Now we can fit the data in the model. Once the model is trained, it can give the prediction. As the sample, we need to provide the data or sample in [ samples, timestamp, features] format; we can cross-check for any three sequential temperatures or look like sequential temperatures. I am using [17, 18, 19] as my sequential temperature to predict the next timestamp temperature value.