Research Article

IntelliPortfolio: Intelligent Portfolio for Enhanced Index Tracking Using Clustering and LSTM

Table 6

Results of different selection algorithms on the SSE180 dataset.

IntelliPortfolio

Selection algorithms
IntelliPortfolio0.005540.001161.765070.24298
Random0.002000.000020.448140.01141
Industry0.00215−0.001200.87021−0.55862
Volume0.00171−0.000300.62009−0.17336
Autoencoder0.004750.000551.265640.09968

GA-RRL
Selection algorithms
IntelliPortfolio0.011030.000940.586420.11496
Random0.005100.000500.244460.09727
Industry0.00546−0.001440.35697−0.26416
Volume0.004410.000110.354480.02391
Autoencoder0.008180.000840.522520.07605

DDPG
Selection algorithms
IntelliPortfolio0.024310.000470.162690.01950
Random0.003650.000120.078260.03227
Industry0.00581−0.000170.05993−0.02978
Volume0.00296−0.00006−0.23251−0.02252
Autoencoder0.017850.00021−0.170500.01179

RRL
Selection algorithms
IntelliPortfolio0.040020.001170.022980.01953
Random0.01092−0.00011−0.01936−0.01017
Industry0.01738−0.00008−0.03844−0.00483
Volume0.010290.00019−0.049260.01857
Autoencoder0.045930.000900.020320.00753

DPG
Selection algorithms
IntelliPortfolio0.002660.00118−0.037300.36666
Random0.00179−0.00024−0.14548−0.13338
Industry0.00085−0.00018−0.36048−0.21024
Volume0.001410.00008−0.171270.05529
Autoencoder0.002250.00060−0.051510.26669