优胜方案合集

汇集各类比赛的优胜方案 · 共 354 个比赛

Abstraction and Reasoning Challenge

2020-02-13 数学与计算组合优化约束满足算法竞技赛

9 个方案

1st place solution + code and official documentation

2nd place solution

My part of the 3rd place solution

5th place short notes

8th Place Solution (Part 1) - Object detection

9th place solution(s)

Our 10th place solution [with code]

20th place solution

21st place solution

Flower Classification with TPUs

2020-02-10 计算机视觉图像分类细粒度识别算法竞技赛

1 个方案

Current first place solution write-up

Categorical Feature Encoding Challenge II

2019-12-31 数值预测文本分类算法竞技赛

2 个方案

1st Place Solution

2nd-Place-solution [xDeepFM Yes!]

Santa 2019 - Revenge of the Accountants

2019-12-23 物流与供应链组合优化约束满足算法竞技赛

1 个方案

Summary: how we found the optimal (?) solution

Bengali.AI Handwritten Grapheme Classification

2019-12-19 计算机视觉图像分类细粒度识别算法竞技赛

18 个方案

1st place solution /w code

2nd place solution

3rd place solution

4th place solution

5th place solution

6th place solution

7th Place Data-based Generalization Solution

8th Place Solution /w Code (minimal ver.)

10th Place: Single Model Using Snapshot Ensemble w/ code

Gold medal - 12 place solution

13th place brief writeup (public 107th)

21st Solution using only Kaggle Kernels

22 place Solution

24th place solution - Post processing(+0.013 Private LB)

38th Place - ZoneMix - A domain specific augmentation

Solution 38, some lessons learned

55th place solution - A frugal approach

Deepfake Detection Challenge

2019-12-11 AI安全与对抗异常检测细粒度识别算法竞技赛

10 个方案

1st place solution

Public 2nd (Private 30th) Place Solution

3rd place solution and code

5th (from 20th) place approach

12th place in a nutshell

27th place solution

39th Private LB: small and simple

43rd place private LB solution

Silver zone solo, no credits used.

Public 34th/Private 47th place insights - Part1 (Inference)

Santa's Workshop Tour 2019

2019-11-27 物流与供应链组合优化约束满足算法竞技赛

5 个方案

How to win Santa's Workshop Tour

Summary: how we found the optimal (?) solution

John does California Odyssey (with code)

My optimal solution

Google QUEST Q&A Labeling

2019-11-22 自然语言处理文本分类文本匹配算法竞技赛

29 个方案

1st place solution with code

Two BERTs are better than one (2nd place solution)

3rd place solution

4th place solution overview

6th place solution

7th place post-processing and ensembling

9th public and private solution: 30 models, 1 for each column?

Private LB 10th Place Solution

11th Solution: A threshold oriented Competition is not good

12th place post-processing

How We Found "Magic" (13th Solution Overview)

15th solution with codes

16th Place Solution

18th place solution (hypotheses on the datasets)

21st place solution

22nd place solution(NLP beginner)

23th place solution

24th place - mixing all I can - BERT+XLNET+GPT2+LGB

26 place solution

30th Sliver Solution (Post-Processing Magic Discovered)

39th place solution, submission errors and memory leakage

41-th Place Solution

silver solution (44th)

45th Place Simple Solution - 4 model average

47th place solution

A general walk through my solution with trick 55th

TensorFlow 2.0 Question Answering

2019-10-28 自然语言处理阅读理解开放域问答算法竞技赛

17 个方案

1st place solution with code

2nd place solution

3rd place solution

4th place Solution

6th place solution

7th place solution

8th place solution

9th place solution

Brief summary of 13th place solution (hide the pain Harold)

17th Place solution [bert-disjoint] kernel + all utility scripts

21th place solution, puzzlingly shaking from public LB 3th

23rd place solution: ensemble, rank passage and predict span

27th solution with luck and some questions from me

30th Place Solution

31st solution with custom loss

45th solution, my journey and learnings, feeling grateful :)

47th Place Solution Write-Up (Ensembling)

2019 Data Science Bowl

2019-10-24 学习效果预测数值预测时序预测算法竞技赛

19 个方案

1st place solution

2nd place solution

3rd solution - single TRANSFORMER model, link to kernel

4th private (7th public) place writeup, link to code

7th Place Solution

8th place solution

9th place solution

10th place Solution

14th place solution

15th palce solution

16th place solution

19th place solution

20th place solution 😂

21st place solution (link to R kernel)

30th Place Write Up

Single Model - 1.5 Transformers - 31st place

40th place solution 0.553 - 0.555, +1050 positions

Solution and some ideas

44 place writeup(Catboost ranking with eventdata)

上一页
...
31
32
33
34
35
...
下一页