└── README.md /README.md: -------------------------------------------------------------------------------- 1 | This repo. is for MAT(meta-AI Team) in 2nd Dacrew as dacon supporters. 2 | We are studying about __AutoML__ and write some contents in Korean. 3 | 4 | In this page, you can get the contents list of the MAT and the links for the contents. 5 | 6 | ## Contents 7 | 8 | ### :gem: Preview 9 | 10 | - 0편) [인공지능 위의 인공지능? AutoML을 알아보자](https://dacon.io/codeshare/4760?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) 11 | 12 | ### :gem: What is AutoML? 13 | 14 | - 1편) [AutoML이란 무엇일까](https://dacon.io/codeshare/4844?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [MIA-khm] 15 | 16 | - 1-1편) Search Strategy [woojin-heo] 17 | - 1-2편) Performance Estimation Strategies [0525hhgus] 18 | 19 | - 2편) [HPO(Hyper-parameter optimization)란 무엇일까](https://dacon.io/codeshare/4863?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [nsbg] 20 | - 3편) [NAS(Network Archtecture Search)란 무엇일까](https://dacon.io/codeshare/4879?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 21 | 22 | ### :gem: Let's elaborate on AutoML 23 | 24 | - HPO(Hyper-parameter Optimization) 25 | 26 | - 4-1편) [HPO알아보기: (1) Grid Search](https://dacon.io/codeshare/4922?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 27 | - 4-2편) [HPO알아보기: (2) Random Search](https://dacon.io/codeshare/5122?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 28 | - 4-3편) [HPO알아보기: (3) Bayesian Search](https://dacon.io/codeshare/5125?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 29 | - Simple Survey [nsbg] 30 | 31 | - 4-4-1편) [HPO(PSO) 논문 읽어보기](https://dacon.io/codeshare/5187?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 32 | > A novel LS-SVMs hyper-parameter selection based on particle swarm optimization 33 | - 4-4-2편) [HPO(uDeas) 논문 읽어보기](https://dacon.io/codeshare/5189?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [nsbg] 34 | > Hyperparameter optimization of deep neural network using univariate dynamic encoding algorithm for searches 35 | - 4-4-3편) [HPO(강화학습) 논문 읽어보기](https://dacon.io/codeshare/5190?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 36 | > Efficient hyperparameter optimization through model-based reinforcement learning 37 | - 4-4-4편) [HPO(Population-based Bandit Optimization) 논문 읽어보기](https://dacon.io/codeshare/5195?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 38 | > Provable Efficient Online hyperparameter optimization with population-based bandits 39 | 40 | - NAS(Network Archtecture Search) 41 | 42 | - 5-1편) [NAS알아보기: (1) Evolutionary Algorithm based Method](https://dacon.io/codeshare/5196?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 43 | - 5-2편) [NAS알아보기: (2) Reinforcement Learning based Method](https://dacon.io/codeshare/5198?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 44 | - 5-3편) [NAS알아보기: (3) Gradient Descent based Method](https://dacon.io/codeshare/5199?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 45 | - 5-4편) Simple Survey [nsbg] 46 | 47 | - 5-4-1편) [NAS(확률기반) 논문 읽어보기](https://dacon.io/codeshare/5202?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 48 | > Probabilistic Neural Architecture Search 49 | - 5-4-2편) [NAS(Evolutionary) 논문 읽어보기](https://dacon.io/codeshare/5205?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 50 | > CARS: Continuous Evolution for Efficient Neural Architecture Search 51 | - 5-4-3편) [NAS(Batch Normalization) 논문 읽어보기](https://dacon.io/codeshare/5200?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [nsbg] 52 | > BN-NAS: Neural Architecture Search with Batch Normalization 53 | 54 | ### :gem: Experiments using AutoML Library 55 | 56 | 57 | 1. [Breast Cancer Wisconsin](https://www.kaggle.com/datasets/uciml/breast-cancer-wisconsin-data) 58 | 59 | - [실전 AutoML 1탄-PyCaret 사용해보기](https://dacon.io/codeshare/5161?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 60 | - [실전 AutoML 1탄-Auto-Sklearn 사용해보기](https://dacon.io/codeshare/5160?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 61 | - [실전 AutoML 1탄-AutoKeras 사용해보기](https://dacon.io/codeshare/5126?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [nsbg] 62 | - [실전 AutoML 1탄-FLAML 사용해보기](https://dacon.io/codeshare/5162?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 63 | 64 | 2. [Used car price precition](https://dacon.io/competitions/official/235901/overview/description) 65 | - [실전 AutoML 2탄-EDA](https://dacon.io/competitions/official/235901/codeshare/5097?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) 66 | - [실전 AutoML 2탄: Pycaret](https://dacon.io/competitions/official/235901/codeshare/5191?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [0525hhgus] 67 | - [실전 AutoML 2탄: AutoSklearn](https://dacon.io/competitions/official/235901/codeshare/5193?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [woojin-heo] 68 | - [실전 AutoML 2탄:AutoKeras](https://dacon.io/competitions/official/235901/codeshare/5192?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [nsbg] 69 | - [실전 AutoML 2탄:FLAML](https://dacon.io/competitions/official/235901/codeshare/5194?utm_source=dacrew&utm_medium=197941&utm_campaign=dacrew_2) [mihosang] 70 | 71 | 72 | --- 73 | 74 | - Written and edited: 2022.03~2022.06 75 | 76 | :seedling: Leader 77 | 78 | - :woman_pilot: [MIA-khm](https://github.com/MIA-khm) : [DACON](https://dacon.io/myprofile/197941/home) 79 | 80 | :seedling: Members 81 | 82 | - :man_student: [mihosang](https://github.com/mihosang) : [DACON](https://dacon.io/myprofile/63929/home) 83 | - :woman_student: [0525hhgus](https://github.com/0525hhgus) : [DACON](https://dacon.io/myprofile/315471/home) 84 | - :woman_student: [nsbg](https://github.com/nsbg) : [DACON](https://dacon.io/myprofile/407317/home) 85 | - :woman_student: [woojin-heo](https://github.com/woojin-heo) : [DACON](https://dacon.io/myprofile/424584/home) 86 | 87 | --------------------------------------------------------------------------------