Prediction of multi assets

Generic
1

hello,

below i have shared code of prediction of multiasset prediction using LSTM

i am  facing problem :

1.   i have predicted prices based on open prices iwant to know wheather is it feasible or  not

2. the model  is not able  to take till today prices ,it is taking open prices fron may

         Date        Open
0  2022-05-31  954.187012
1  2022-06-01  956.628601
2  2022-06-02  958.211975
3  2022-06-03  959.224670
4  2022-06-04  959.729797
5  2022-06-05  959.601318
6  2022-06-06  958.860107
7  2022-06-07  957.628906
8  2022-06-08  956.460693
9  2022-06-09  955.743652
10 2022-06-10  956.163513
11 2022-06-11  957.715332
12 2022-06-12  959.994019
13 2022-06-13  962.220154
14 2022-06-14  963.143799
15 2022-06-15  962.100464
16 2022-06-16  958.928284
17 2022-06-17  953.975708
18 2022-06-18  947.488831
19 2022-06-19  939.928040
20 2022-06-20  931.602295
21 2022-06-21  923.777344
22 2022-06-22  916.554688
23 2022-06-23  910.107300
24 2022-06-24  904.733337
25 2022-06-25  900.756958
26 2022-06-26  898.112610
27 2022-06-27  896.669678
28 2022-06-28  896.284302
29 2022-06-29  896.980469

like this but if i run today it should take till today and should predict  next day price.

3. i want to predict price for 15 or 30 days which i am not able to and i wanted all 15days originalprice and predicted prices
  and also i want to find the accuracy or prediction 

please help me as soon as possible.

 
2

Hello Keerti,



Can you please share the code?



Thank you

 

3

please send your mail id will send

 

4

Hi Keerti, thanks for sharing the code, received it from the support team. Will check it and get back to you.

 

5

Hi Keerti,



To answer your questions:


  1. Yes, it is feasible. However, it is a good practice to account for slippage while doing so. You can learn more about the concept of slippage here.


  2. As per my understanding of your explanation, the data you shared above is the input data. The dates in the data look continuous, which may not be possible for stocks as there are market holidays. I think resolving this would resolve your issue.


  3. This question seems a bit ambiguous to me. It could be interpreted in two different ways:

    a.) If you are trying to predict the price for the next 15 days, then you can do so with a loop as you are doing currently.

    b.) However, if you are trying to predict the price on the 15th day by considering today as Day 0, then you can simply change the target variable while training the model to price on Day 15.



    The above answers are as per my understanding of the issue. If otherwise, please feel free to reply to this community post, and we can discuss this further. Hope this helps!



    Thanks,

    Akshay

     
6

if you dont mind Akshay i really need help interms of code. i am in last stage to submit. i am not really  understanding hoe to do this.