Skip to main content

Advertisement

Table 3 Classification results with ANN/DNN classifiers using transformed data with PCs = 60

From: Predicting the daily return direction of the stock market using hybrid machine learning algorithms

# of hidden layers training (MSE) validation (MSE) testing (MSE) total (MSE)
10 58.2 (0.3062) 54.1 (0.3110) 57.8 (0.3091) 57.5 (0.3074)
12 56.9 (0.3079) 53.3 (0.3137) 58.1 (0.3066) 56.6 (0.3086)
14 57.9 (0.3041) 54.6 (0.3135) 57.8 (0.3084) 57.4 (0.3062)
16 59.4 (0.3020) 55.4 (0.3128) 59.9 (0.3056) 58.9 (0.3042)
18 56.7 (0.3071) 54.6 (0.3109) 58.9 (0.3089) 56.7 (0.3080)
20 58.8 (0.3052) 54.4 (0.3109) 59.2 (0.3074) 58.2 (0.3064)
22 57.3 (0.3065) 55.4 (0.3133) 59.4 (0.3083) 57.3 (0.3078)
24 56.9 (0.3080) 54.9 (0.3099) 58.4 (0.3082) 56.8 (0.3083)
26 55.9 (0.3101) 56.0 (0.3105) 58.4 (0.3088) 56.3 (0.3099)
28 57.8 (0.3057) 56.5 (0.3105) 59.4 (0.3079) 57.9 (0.3067)
30 56.2 (0.3076) 53.6 (0.3152) 58.1 (0.3104) 56.1 (0.3092)
35 56.6 (0.3066) 56.2 (0.3134) 58.1 (0.3081) 56.8 (0.3078)
40 59.8 (0.2999) 54.9 (0.3125) 57.6 (0.3095) 58.7 (0.3032)
45 56.3 (0.3096) 54.6 (0.3163) 57.3 (0.3113) 56.2 (0.3109)
50 55.2 (0.3103) 53.6 (0.3154) 57.3 (0.3078) 55.3 (0.3107)
100 56.9 (0.3077) 53.1 (0.3205) 57.6 (0.3221) 56.4 (0.3117)
500 55.5 (0.3345) 54.9 (0.3309) 59.9 (0.3162) 56.1 (0.3312)
1000 58.4 (0.3240) 55.7 (0.3392) 58.1 (0.3285) 57.9 (0.3269)