# Introduction to Machine Learning
###### tags: `MLCC`

## 通過這個課程可以學到3件事
1. 縮短編程時間
- 有許多現成的工具，你只需要提供一些資料樣本就可以開發程式
3. 客製化自己的產品/程式
- 當已經有一個機器學習模型，就可以套用到其他n個程式，改變依你的需求與資料而定
3. 解決不知如何用人工解決的問題
- 對於太複雜的問題，只要向機器提供大量樣本，機器就會自己找出解決方案
4. 改變思考方式
- 我們可以用機器學習，透過實驗和統計，觀察這個充滿不確定性的未知世界，而不是以邏輯線性地推導理論
## 常用術語

| 名詞 | 解釋 |
| -------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Instance | The thing about which you want to make a prediction |
| **Label** | An answer for a prediction task, or which is produced by a ML system, or the right answer in training data |
| **Feature** | A property of an instance used in a prediction task |
| Feature Column | A set of related features |
| Example | An instance (with its features) and a label |
| Model | A statistical representation of a prediction task |
| **Metric** | A number that you care about |
| Objective | A metric that your algorithm is trying to optimize |
| Pipeline | The infrastructure surrounding a machine learning algorithm.(Includes gathering the data and training the models) |

## [Rules of Machine Learning](https://developers.google.com/machine-learning/guides/rules-of-ml#rule_1_dont_be_afraid_to_launch_a_product_without_machine_learning)

## 學習機器學習的前置準備(?)
1. 確保Pipeline是可靠、穩定的
2. 從「合理的目標」開始做
3. 找 common sense feature
> common sense 指的是「大家都知道的道理/常理」，通常是依據**直覺或人情事理**的判斷