Cancer Risk Prediction on EHR-based cancer screening dataset 2016-2023 (test)
22.8Average PrecisionCan-SAVE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Can-SAVEPipeline=Proposed Method2023.09 | 22.8 | 83.7 | |
| DeepSeek-R1 -> Qwen3-Emb -> LoRAPipeline=Supervised Fine-Tuned LLM (LoRA) Pipeline2023.09 | 19.3 | 90.1 | |
| Qwen3-Emb ->DeepHitPipeline=LLM Encoding Pipeline2023.09 | 18.6 | 88.5 | |
| GigaChat ->GBMPipeline=LLM Encoding Pipeline2023.09 | 18.5 | 89.6 | |
| DeepSeek-R1 -> Qwen3-Emb -> GBMPipeline=LLM Summarization & Encoding Pipeline2023.09 | 17.6 | 88.1 | |
| DeepSeek-R1 -> Qwen3-Emb -> DeepHitPipeline=LLM Summarization & Encoding Pipeline2023.09 | 17.4 | 89.5 | |
| DeepSeek-R1 ->GBMPipeline=LLM Encoding Pipeline2023.09 | 16.4 | 87.3 | |
| GBM (CatBoost)Pipeline=Machine Learning Pipeline2023.09 | 16 | 78.6 | |
| BERT ->GRUPipeline=Deep Learning (RNN & Transformers) Pipeline2023.09 | 15.1 | 84.9 | |
| Qwen3-Emb ->GBMPipeline=LLM Encoding Pipeline2023.09 | 15.1 | 86.9 | |
| AFT modelPipeline=Survival Models Pipeline2023.09 | 11.7 | 84.8 | |
| Logistic RegressionPipeline=Machine Learning Pipeline2023.09 | 10.4 | 83.4 | |
| Fine-tuned CoLESPipeline=Deep Learning (RNN & Transformers) Pipeline2023.09 | 10.3 | 81.3 | |
| Randon ForestPipeline=Machine Learning Pipeline2023.09 | 10.2 | 83.3 | |
| DeepHitPipeline=Survival Models Pipeline2023.09 | 10.2 | 86.4 | |
| Deep Survival MachinesPipeline=Survival Models Pipeline2023.09 | 10.1 | 82.3 | |
| Longformer ->GBMPipeline=Deep Learning (RNN & Transformers) Pipeline2023.09 | 9.3 | 77.7 | |
| Randon Survival ForestPipeline=Survival Models Pipeline2023.09 | 7.4 | 78.6 |