Machine-Learning

This is a story related to machine learning and all data science skill.

Keys of Machine Learning

• Steps for building a Machine Learning

1.Asking Questions 2.Data Preparation(Preprocessing:Categorial feature,missing value and standardization;feature selection) 3.Building Models with Different Hyperparameters 4.Test and Evaluate Models

• Machine Learning Models

1.Objective function (loss function) 2.Approximation function (to approximate the true function) 3.Optimization Method (to optimize the objective function)

Machine Learning Roadmap

| Name | iThome 鐵人賽 | Material & Assignment & Reference | | -------- | -------- | -------- | | Level 1 : Python Basic Skills | Day 01：Python 介紹與開發環境
Day 02：Python 基礎觀念 (1)
Day 03：Python 基礎觀念 (2)
Day 04：Python 基礎觀念 (3)| A：01H1Basic
A：01S1Basic
R：Codecadedy Learn Python 3| | Level 2 : Data Preprocessing | Day 05：Pandas 操作 (1)
Day 06：Pandas 操作 (2) | A：02H1Pandas
A：02S1Pandas
R：numpy和pandas中 axis(軸)概念
R：Excel與Pandas之間的愛恨糾葛1
R：Excel與Pandas之間的愛恨糾葛2
R：Excel與Pandas之間的愛恨糾葛3
R：Numpy & Pandas 簡介| |Level 3 : Data Visualizing|Day 07：Matplotlib 操作
Day 08：Seaborn 操作 | A：03H1Visualizing
A：03S1Visualizing | |Level 4 : Introduction of Tools||[R：[Day02]Jupyter Notebook操作](https://ithelp.ithome.com.tw/articles/10192614)
R：Jupyter Notebook介紹及安裝
R：JupyterLab
R：Vscode
R：Hackmd 常用 LaTeX| |Level 5 : Database Related|Day 09：資料庫介紹
Day 10：Postgres 操作
Day 11：psycopg2 操作|M：05M2Postgresandpsycopg2| |Level 6 : Python Advanced Skills| Day 12：物件導向
Day 13：程式除錯與異常
Day 14：程式碼日誌與品質| M：06M1Object-Oriented-Programming
M：06M2ErrorandException
M：06M3CleanCode
M：06M4_Decorator| |Level 7 : Model Prerequisite Knowledge|Day 15：機器學習介紹
Day 16：模型衡量指標
Day 17：資料預處理 (1)
Day 18：資料預處理 (2)
|M：07M1ModelPrerequisiteKnowledge
M：07M2DataPreprocessing
[R：[Day24]什麼是機器學習？](https://ithelp.ithome.com.tw/articles/10196922)
R：李宏毅教授的影片
Introduction of Machine Learning
Regression-Case Study
Regression-Demo
What does the error come from?
Gradient Descent
Classification
Logistic Regression| |Level 8 : Model Development|Day 19：KNN 與 K-means
Day 20：線性迴歸與羅吉斯迴歸
Day 21：SVM
Day 22：決策樹
Day 23：集成式學習
Day 24：隨機森林
Day 25：XGBoost
Day 26：LightGBM 與 GridSearch
Day 27：模型解釋 Shap
|M：08M0BackgroundKnowledge
M：08M1KNN&K-Means
M：08M2Regression
M：08M3SVM
M：08M4DecisionTree
M：08M5RandomForest
M：08M6XgBoost
M：08M7LightGBM
M：08M8GridSearch
M：08M9Shap| |Level 9 : Git Tutorial|Day 28：Git|M：09M1Git | |Level 10 : API Service|Day 29：FastAPI 讓模型上線| M：10APIService| |Level 11 : Model Monitoring|||

Advanced

| Name | Material & Assignment & Reference | | -------- | -------- | | Level 12 : Deep Learning|M：TF-IDF使用
M：nltk
M：What's cooking
M：Keras PDF
M：Keras Demo
R：卷積神經網絡介紹(Convolutional Neural Network)
R：關於影像辨識，所有你應該知道的深度學習模型
R：OCR技术：大批量构造中文文字训练集
R：Google Cloud Vision API
R：Google Cloud Text API|