Questions about experimental result

[aixiaodewugege]

No description provided.

[M-Nauta]

The results in Table 4 of the paper are the average over all datasets in the Finance and fMRI folders. I just ran the following commands again with the default parameters (except for the number of epochs which is set to 5000 as also mentioned in the paper) and got similar results as reported in the paper (even higher F1 scores for fMRI). Note that results might differ slightly with different random seeds and due to some randomness in PyTorch executions.

Finance:
runTCDF.py --cuda --epochs 5000 --ground_truth data/Finance/random-rels_20_1A_returns30007000_header.csv=data/Finance/random-rels_20_1A.csv,data/Finance/random-rels_20_1B_returns30007000_header.csv=data/Finance/random-rels_20_1B.csv,data/Finance/random-rels_20_1C_returns30007000_header.csv=data/Finance/random-rels_20_1C.csv,data/Finance/random-rels_20_1D_returns30007000_header.csv=data/Finance/random-rels_20_1D.csv,data/Finance/random-rels_20_1E_returns30007000_header.csv=data/Finance/random-rels_20_1E.csv,data/Finance/random-rels_20_1_3_returns30007000_header.csv=data/Finance/random-rels_20_1_3.csv,data/Finance/random-rels_40_1_3_returns30007000_header.csv=data/Finance/random-rels_40_1_3.csv,data/Finance/random-rels_40_1_returns30007000_header.csv=data/Finance/random-rels_40_1.csv,data/Finance/manyinputs_returns30007000_header.csv=data/Finance/manyinputs.csv

fMRI:
runTCDF.py --cuda --epochs 5000 --ground_truth data/fMRI/timeseries1.csv=data/fMRI/sim1_gt_processed.csv,data/fMRI/timeseries2.csv=data/fMRI/sim2_gt_processed.csv,data/fMRI/timeseries3.csv=data/fMRI/sim3_gt_processed.csv,data/fMRI/timeseries5.csv=data/fMRI/sim5_gt_processed.csv,data/fMRI/timeseries6.csv=data/fMRI/sim6_gt_processed.csv,data/fMRI/timeseries7.csv=data/fMRI/sim7_gt_processed.csv,data/fMRI/timeseries8.csv=data/fMRI/sim8_gt_processed.csv,data/fMRI/timeseries9.csv=data/fMRI/sim9_gt_processed.csv,data/fMRI/timeseries10.csv=data/fMRI/sim10_gt_processed.csv,data/fMRI/timeseries11.csv=data/fMRI/sim11_gt_processed.csv,data/fMRI/timeseries12.csv=data/fMRI/sim12_gt_processed.csv,data/fMRI/timeseries13.csv=data/fMRI/sim13_gt_processed.csv,data/fMRI/timeseries14.csv=data/fMRI/sim14_gt_processed.csv,data/fMRI/timeseries15.csv=data/fMRI/sim15_gt_processed.csv,data/fMRI/timeseries16.csv=data/fMRI/sim16_gt_processed.csv,data/fMRI/timeseries17.csv=data/fMRI/sim17_gt_processed.csv,data/fMRI/timeseries18.csv=data/fMRI/sim18_gt_processed.csv,data/fMRI/timeseries19.csv=data/fMRI/sim19_gt_processed.csv,data/fMRI/timeseries20.csv=data/fMRI/sim20_gt_processed.csv,data/fMRI/timeseries21.csv=data/fMRI/sim21_gt_processed.csv,data/fMRI/timeseries22.csv=data/fMRI/sim22_gt_processed.csv,data/fMRI/timeseries23.csv=data/fMRI/sim23_gt_processed.csv,data/fMRI/timeseries24.csv=data/fMRI/sim24_gt_processed.csv,data/fMRI/timeseries25.csv=data/fMRI/sim25_gt_processed.csv,data/fMRI/timeseries26.csv=data/fMRI/sim26_gt_processed.csv,data/fMRI/timeseries27.csv=data/fMRI/sim27_gt_processed.csv,data/fMRI/timeseries28.csv=data/fMRI/sim28_gt_processed.csv