機器學習100天挑戰

Day 25 時間特徵

透過時間運算來處理頻率問題

Day 26 特徵組合

透過數學運算來修正特徵(EX: 兩點間的距離考慮地球的經緯度導致的距離差)

Day 27 群聚編碼 (Group by Encoding)

使用 pandas 的GroupBy( )來做特徵處理

Day 28 相關係數

Day 29 特徵評估

藉由sklearn的GradientBoostingClassifier的feature_importances_來分配特徵重要性

Day 30 ROC Curve

Day 31 機器學習概論(影片)

Day 32 機器學習文章

找一篇有興趣的paper/ article

Day 33 機器學習流程(影片)

Day 34 切割訓練&測試集

學習切割分佈不平均的資料

Day 35 Multi Label Case

Day 36 評估指標

評估Regression與Classification的準確度

Day 37 Regression介紹文

Day 38 Linear Regression實作

Day 39 Lasso & Ridge

Linear Regression加上L1/ L2 Regression後

Day 40 Linear vs Lasso vs Ridge

Day 41 Decision Tree概念

Day 42 Decision Tree實作Regression/ Classification

Day 43 理解Random Forest

假設總共有 N 筆資料，每棵樹用取後放回的方式抽了總共 N 筆資料生成，請問這棵樹大約使用了多少 % 不重複的原資料生成?

Day 44 Random Forest Classification/ Regression

Day 45 Boosting

Bagging: 減少overfitting的機會
Boosting: 處理underfitting，增加performance

Day 46 Gradeint Boosting Tree

透過sklearn實作

Day 47 超參數

sklearn gridSearch學找最佳參數組合

Day 54 & 55 K-means

sklearn kmeans & plot clusters

Day 56 K-means Coding

sklearn silhouette評估分群績效 & silhouette作圖

Day57 Hierarchical Clustering

使用sklearn的hierarchical clustering演算法

Day 59 PCA的基本認識

Day 60 Sklearn PCA

研究sklearn PCA & pipeline
研究多個axis跨cell不能作圖

Day 61 t-SNE

用手寫數字跑t-SNE

Day 62 t-SNE作圖

Day 66 安裝keras

了解keras的backend functions

Name		Name	Last commit message	Last commit date
Latest commit History 65 Commits
Day49toDay53		Day49toDay53
data		data
Day16.png		Day16.png
Day_001_HW.ipynb		Day_001_HW.ipynb
Day_002_HW.ipynb		Day_002_HW.ipynb
Day_003-1_HW.ipynb		Day_003-1_HW.ipynb
Day_003-2_HW.ipynb		Day_003-2_HW.ipynb
Day_004_HW.ipynb		Day_004_HW.ipynb
Day_005_HW.ipynb		Day_005_HW.ipynb
Day_006_HW.ipynb		Day_006_HW.ipynb
Day_006_outliers_detection_ANS.ipynb		Day_006_outliers_detection_ANS.ipynb
Day_007_HW.ipynb		Day_007_HW.ipynb
Day_008_HW.ipynb		Day_008_HW.ipynb
Day_010_HW.ipynb		Day_010_HW.ipynb
Day_011_HW.ipynb		Day_011_HW.ipynb
Day_013_HW.ipynb		Day_013_HW.ipynb
Day_014_HW.ipynb		Day_014_HW.ipynb
Day_015_HW.ipynb		Day_015_HW.ipynb
Day_016_HW.ipynb		Day_016_HW.ipynb
Day_017_HW.ipynb		Day_017_HW.ipynb
Day_018_HW.ipynb		Day_018_HW.ipynb
Day_019_HW.ipynb		Day_019_HW.ipynb
Day_020_HW.ipynb		Day_020_HW.ipynb
Day_021_HW.ipynb		Day_021_HW.ipynb
Day_022_HW.ipynb		Day_022_HW.ipynb
Day_023_HW.ipynb		Day_023_HW.ipynb
Day_024_HW.ipynb		Day_024_HW.ipynb
Day_025_HW.ipynb		Day_025_HW.ipynb
Day_026_HW.ipynb		Day_026_HW.ipynb
Day_027_HW.ipynb		Day_027_HW.ipynb
Day_028_HW.ipynb		Day_028_HW.ipynb
Day_029_HW.ipynb		Day_029_HW.ipynb
Day_030_HW.ipynb		Day_030_HW.ipynb
Day_031_HW.ipynb		Day_031_HW.ipynb
Day_032_HW.ipynb		Day_032_HW.ipynb
Day_033_HW.ipynb		Day_033_HW.ipynb
Day_034_HW.ipynb		Day_034_HW.ipynb
Day_035_HW.ipynb		Day_035_HW.ipynb
Day_036_HW.ipynb		Day_036_HW.ipynb
Day_037_HW.ipynb		Day_037_HW.ipynb
Day_038_HW.ipynb		Day_038_HW.ipynb
Day_039_HW.ipynb		Day_039_HW.ipynb
Day_040_HW.ipynb		Day_040_HW.ipynb
Day_041_HW.ipynb		Day_041_HW.ipynb
Day_042_HW.ipynb		Day_042_HW.ipynb
Day_043_HW.ipynb		Day_043_HW.ipynb
Day_044_HW.ipynb		Day_044_HW.ipynb
Day_046_HW.ipynb		Day_046_HW.ipynb
Day_047_HW.ipynb		Day_047_HW.ipynb
Day_054_HW.ipynb		Day_054_HW.ipynb
Day_055_HW.ipynb		Day_055_HW.ipynb
Day_056_HW.ipynb		Day_056_HW.ipynb
Day_057_HW.ipynb		Day_057_HW.ipynb
Day_059_HW.ipynb		Day_059_HW.ipynb
Day_060_HW.ipynb		Day_060_HW.ipynb
Day_061_HW.ipynb		Day_061_HW.ipynb
Day_062_HW.ipynb		Day_062_HW.ipynb
Day_066_HW.ipynb		Day_066_HW.ipynb
Day_068_HW.ipynb		Day_068_HW.ipynb
Day_070_HW.ipynb		Day_070_HW.ipynb
Day_071_HW.ipynb		Day_071_HW.ipynb
README.md		README.md

johnny12150/100Day-ML-Marathon

Folders and files

Latest commit

History

Repository files navigation