Skip to main content

Table 5 Validation accuracy of the LSTM and GRU architectural models

From: Implementation of Long Short-Term Memory and Gated Recurrent Units on grouped time-series data to predict stock prices accurately

Company

Model

Upper accucacy

(MAPE)

Average accuracy

(RMSPE)

Lower accuracy

(RMDPE)

LSTM

GRU

LSTM

GRU

LSTM

GRU

AMZN

Model 1

96.06%

94.23%

94.91%

93.31%

85.06%

82.62%

Model 2

94.19%

95.48%

92.50%

94.63%

76.41%

87.19%

Model 3

93.92%

95.16%

92.22%

94.02%

76.02%

82.54%

Model 4

88.34%

96.85%

86.53%

96.07%

67.89%

87.19%

GOOGL

Model 1

89.12%

93.65%

86.78%

92.64%

75.10%

81.66%

Model 2

92.01%

95.44%

90.31%

94.33%

81.05%

85.43%

Model 3

90.56%

92.14%

88.09%

89.78%

77.13%

79.18%

Model 4

88.97%

93.87%

87.36%

92.65%

75.49%

80.27%

BLL

Model 1

93.26%

97.37%

92.61%

96.60%

86.48%

84.98%

Model 2

91.31%

95.63%

90.36%

95.05%

82.93%

87.14%

Model 3

88.12%

91.46%

87.51%

90.97%

80.79%

85.95%

Model 4

95.60%

97.15%

94.46%

96.30%

86.26%

82.27%

QCOM

Model 1

94.46%

95.11%

93.51%

94.47%

82.45%

84.56%

Model 2

90.58%

94.21%

89.67%

93.60%

76.20%

83.69%

Model 3

89.37%

96.14%

88.38%

95.38%

74.61%

84.62%

Model 4

93.72%

95.82%

92.42%

95.13%

79.63%

85.55%