YalaLab

/

Pillar0-Sybil-1.5

like 3

Follow

Yala Lab 45

Thank you for sharing your excellent work!

I applied your repository exactly as it is to the CT-RATE and NLST dataset. However, I noticed something strange and had a question. The predicted lung cancer risk values for the 6-year period are outputting almost identical logit values for all 6 years within each sample.

Here are the results from ensemble.csv:

[accession, survival]
valid_1_a_1,"[0.0065, 0.0065, 0.0065, 0.0065, 0.0066, 0.0066]"
valid_1_a_2,"[0.0165, 0.0165, 0.0165, 0.0165, 0.0165, 0.0165]"
valid_2_a_1,"[0.1727, 0.1727, 0.1727, 0.1727, 0.1727, 0.1727]"
valid_2_a_2,"[0.1360, 0.1360, 0.1360, 0.1360, 0.1360, 0.1410]"
valid_3_a_1,"[0.0366, 0.0366, 0.0366, 0.0366, 0.0366, 0.0366]"
valid_3_a_2,"[0.0923, 0.0923, 0.0923, 0.0923, 0.0923, 0.0923]"
valid_4_a_1,"[0.2736, 0.2736, 0.2736, 0.2736, 0.2736, 0.2736]"
valid_4_a_2,"[0.5117, 0.5117, 0.5117, 0.5117, 0.5117, 0.5117]"
valid_4_b_1,"[0.1707, 0.1707, 0.1707, 0.1707, 0.1707, 0.1707]"
valid_4_b_2,"[0.2178, 0.2178, 0.2178, 0.2178, 0.2178, 0.2178]"
valid_5_a_1,"[0.0079, 0.0079, 0.0079, 0.0079, 0.0079, 0.0079]"
valid_5_a_2,"[0.0131, 0.0131, 0.0131, 0.0131, 0.0131, 0.0131]"
valid_6_a_1,"[0.0047, 0.0047, 0.0047, 0.0047, 0.0047, 0.0047]"
valid_6_a_2,"[0.0071, 0.0071, 0.0071, 0.0071, 0.0071, 0.0071]"
valid_7_a_1,"[0.0506, 0.0506, 0.0506, 0.0506, 0.0506, 0.0506]"