# Is it possible to get prediction intervals in sequenced data RNN forecasting?

Is it possible to get prediction intervals in sequenced data RNN (keras+python)forecasting?

For example: predicts your car sales or new purchase

The question is: UserId 1 will change his car in the next 2 years?

Uid BUY                 SELL                BRAND USEDorNEW
1   2013-11-11 00:00:00 2099-12-31 00:00:00 Ford  U
1   2015-04-09 00:00:00 2015-04-13 00:00:00 Opel  U
2   2012-05-10 00:00:00 2019-07-04 00:00:00 Opel  N
2   2019-07-16 00:00:00 2099-12-31 00:00:00 WV    U
3   2009-12-29 00:00:00 2010-12-27 00:00:00 Ford  U
3   2011-01-04 00:00:00 2012-04-18 00:00:00 WV    U
3   2011-05-19 00:00:00 2015-01-28 00:00:00 Ford  U
3   2012-09-28 00:00:00 2015-01-04 00:00:00 WV    N
3   2015-01-02 00:00:00 2099-12-31 00:00:00 WV    U
3   2015-02-18 00:00:00 2099-12-31 00:00:00 WV    N
4   2013-05-13 00:00:00 2018-06-19 00:00:00 AUDI  U
4   2018-06-21 00:00:00 2099-12-31 00:00:00 Ford  U
5   2015-05-05 00:00:00 2019-05-13 00:00:00 WV    N
5   2016-05-16 00:00:00 2099-12-31 00:00:00 Opel  U
5   2017-03-01 00:00:00 2099-12-31 00:00:00 Opel  U
5   2019-05-21 00:00:00 2099-12-31 00:00:00 AUDI  U


I have made OHE, scaling, and padding but what is the next step?

Uid TimeToNew Time used   ford opel wv audi
1   na        1048.7      1    0    0  0
1   514       0.133333333 0    1    0  0
2   na        87.03333333 0    1    0  0
2   2623      979.6       0    0    1  0
3   na        12.1        1    0    0  0
3   371       15.66666667 0    0    1  0
3   135       45          1    0    0  0
3   498       27.6        0    0    1  0
3   826       1034.8      0    0    1  0
3   47        1033.233333 0    0    1  0
4   na        62.1        0    0    0  1
4   1865      992.6       1    0    0  0
5   na        48.96666667 0    0    1  0
5   377       1018.133333 0    1    0  0
5   289       1008.5      0    1    0  0
5   811       981.4666667 0    0    0  1


Two important things: I have a much longer sequence for this task. It is only the sample/example. I know how I use RNN (eg. creating model, reshaping data,etc), but in don't know how can I make prediction intervals.