paper-with-me

Text Classification 벤치마크

Text Classification on DODF Data

10개 결과 · ⬇ CSV · JSON

Average F1

0.8374 0.851 0.8646 0.8782 0.8918 2020-03 2026-09 ULMFiT (pre-trained vocab, no gradual unfreezing) — 0.8918 (2020-03-02) SVM + word counts (pre-trained vocab) — 0.8782 (2020-03-02) SVM + tf-idf (no pre-trained vocab) — 0.8755 (2020-03-02) ULMFiT (no pre-trained vocab) — 0.8469 (2020-03-02) ULMFiT (pre-trained vocab) — 0.8374 (2020-03-02) ULMFiT (pre-trained vocab, no gradual unfreezing) — 0.8918 (2020-03-02) SVM + word counts (pre-trained vocab) — 0.8782 (2020-03-02) SVM + tf-idf (no pre-trained vocab) — 0.8755 (2020-03-02) ULMFiT (no pre-trained vocab) — 0.8469 (2020-03-02) ULMFiT (pre-trained vocab) — 0.8374 (2020-03-02) ULMFiT (pre-trained vocab, no gradual unfreezing) — 0.8918 (2020-03-02)
RankModel Average F1Weighted F1 PaperCodeYear
1 ULMFiT (pre-trained vocab, no gradual unfreezing) 0.89180.9257 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
2 SVM + word counts (pre-trained vocab) 0.87820.9049 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
3 SVM + tf-idf (no pre-trained vocab) 0.87550.8917 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
4 ULMFiT (no pre-trained vocab) 0.84690.8974 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
5 ULMFiT (pre-trained vocab) 0.83740.9088 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
6 ULMFiT (pre-trained vocab, no gradual unfreezing) 0.89180.9257 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
7 SVM + word counts (pre-trained vocab) 0.87820.9049 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
8 SVM + tf-idf (no pre-trained vocab) 0.87550.8917 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
9 ULMFiT (no pre-trained vocab) 0.84690.8974 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
10 ULMFiT (pre-trained vocab) 0.83740.9088 Inferring the source of official texts: can SVM beat ULMFiT? peluz/kneedle-exploration 2020
1–10 / 10 페이지당 10 20 50 100