House Sales in King County, USA
===
Predict house price using kNN in R

**[English](https://hackmd.io/s/SJ2hlMQz7)**
**[中文版](https://hackmd.io/s/BJGOjXWRM)**

## DATABASE


FROM : [Kaggle](https://www.kaggle.com/harlfoxem/housesalesprediction)

## Why choose KING COUNTY in Kaggle?

:::info
Ans :
Taiwan government data open platform lacks relevant complete information
And because it is an open source project, I hope to establish a **complete forecasting paradigm**
:::

## Program environment

> **R** language
**LINUX**

:::warning
**Originally wanted to use Rstudio, but after discussing it with the TA, I felt that developing R under Linux would be better than software:**
- R has many packages that are only available on linux
- What this project needs to do is open source project
- I do not know if someone will use this project (?) in the future, so open source under Linux is better:smiley:
- More unknown areas can be explored when building an environment under Linux:100:
:::

## Schedule

gantt
    title A Gantt Diagram
    
    section Build
    build R           :a0 , 2018-05-05, 18d

    section data_preprocessing
    database         :2018-05-13 , 10d
    preprocess         :2018-05-23  , 10d
    
    section Train/Test
    select feature        :2018-06-02, 10d
    
Improve accuracy      :2018-06-12 ,12d

## BULID R on LINUX

reference [Technical documents](https://hackmd.io/s/r1R3MRkgQ)

## preprocess

reference [Technical documents](https://hackmd.io/s/r1R3MRkgQ)

## reference
[R語言進行機器學習方法及實例（一）](https://hk.saowen.com/a/57cac19c62d7cc640dae3471360106c1d0475e09d4c9032621124c9242335c86)

---