Dept. | Student No. | Name |
---|---|---|
資科碩一 | 110753140 | 張立暘 |
資科三 | 108703017 | 邱彥翔 |
資科碩一 | 110753163 | 林昱辰 |
社會二 | 109204035 | 黃楷捷 |
Our goal is to predict the salary of STEM jobs !
You should provide an example commend to reproduce your result
-
We put our EDA & Model result into shiny app
ShinyApps link :
- docs
- Presentation Slide
-
Source
-
Input format
- One .csv file.
- Attribute Information:
-
Any preprocessing?
- drop the NA
- delete the outlier
-
Which method do you use?
- Linear regression
- Decision tree regression
- Random forest regression
- SVM
- XGBtree
-
What is a null model for comparison?
- Guess the median salary
-
How do your perform evaluation?
- Cross-validation
- MAE
-
Which metric do you use
- MAE
-
Is your improvement significant?
- Yes , from 41809.7 to 18185.23
-
What is the challenge part of your project?
- NA值很多,嘗試用KNN來補效果卻不大好
- Shiny app 呈現會有一些大小的問題,以及無法正確visualization
- data science的project分工以及merge code是一個大問題
- corrplot
- caret
- rpart
- ROCR
- e1071
- randomForest
- Formula
- Metrics
- gbm
- ggbiplot
- ggplot2
- sf
- data.table
- tidyverse
- maps
- repr
- ggthemes
- scales
- ggpubr
- shinythemes
- shiny
- shinydashboard
- cowplot
- rgdal
- e1071
- mlbench
- MLmetrics