| Tartalom: |
http://hdl.handle.net/10890/54996 |
| Archívum: |
Műegyetem Digitális Archívum |
| Gyűjtemény: |
1. Tudományos közlemények, publikációk
Konferenciák gyűjteményei
2nd Workshop on Intelligent Infocommunication Networks, Systems and Services, 2024
|
| Cím: |
Selection and LSTM based trading of sparse, optimal portfolios using the VAR(p) model
|
| Létrehozó: |
Rácz, Attila
Fogarasi, Norbert
|
| Dátum: |
2024-02-26T15:42:55Z
2024-02-26T15:42:55Z
2024
|
| Tartalmi leírás: |
Given a multivariate time series of market observable security prices, this paper explores the use of VAR(p) model to select sparse portfolios that maximize predictability and uses LSTM neural network based trading strategies to assess their profitability. Several previous papers have used the VAR(1) model to show the viability of this method for selecting sparse, mean reverting portfolios and have shown its profitability. This work builds on this and shows that higher dimensional VAR(p) models can achieve a better fit of the multivariate time series and therefore result in higher profits using LSTM neural network based convergence trading algorithms.
|
| Nyelv: |
angol
|
| Típus: |
Konferenciaközlemény
|
| Formátum: |
application/pdf
|
| Azonosító: |
|
