Skip to main content
Financial Innovation

Table 7 Precision values PLSTM for the LSTM classification and differences PLSTM − PRC for PRC = 0.49 of a random choice classifier with probability of class 1 p(1) = 0.49

From: Forecasting relative returns for S&P 500 stocks using machine learning

No

Stock

PLSTM

PLSTM − PRC

1

AMAT

0.71

0.22

2

BSX

0.46

− 0.03

3

ENPH

0.83

0.34

4

ETSY

0.73

0.24

5

LNT

0.39

− 0.09

⋮

⋮

⋮

⋮

493

SIVB

0.66

0.17

494

ZTS

0.52

0.03

Mean value

0.503

0.012

Pooled value

0.51

0.02

  1. The LSTM mean precision value PLSTM is larger than that of the random choice classifier (p = 0.008)
Back to article page