Publications by Nguyen Chi Dung
Predicting Employee Churn: A Short Comparision between Xgboost and Keras Deep Learning
Xgboost vs. Keras Deep Learning Tò mò tìm hiểu coi những thuật toán nào thường được sử dụng bởi các đội thuộc Top 5 trên sân chơi Kaggle về khoa học dữ liệu thì vấp phải cụm từ Primary ML software used by top-5 teams on Kaggle với cái hình như trên. Đại ý là Keras Deep Learning xếp �...
19728 sym R (12028 sym/26 pcs) 2 img
Classification Cutoff Maximizes Profit using Bayesian Optimization
To Be Or Not To Be: Profit Xếp hạn tín dụng. Xếp hạn tín dụng (Credit Classificatin hay Credit Scoring) đóng một vai trò quan trọng đối với lợi nhuận và phát triển bền vững của ngân hàng nói riêng cũng như các tổ chức tài chính cho vay khác. Sự kiện cuộc khủng hoảng tài chính dẫn...
36099 sym R (11637 sym/25 pcs) 1 img 4 tbl
Bayesian Optimization for Turning XGBoost Hyperparameter
About Xgboost Xgboost là một thuật toán Machine Learning được phát triển và hình thành từ một dự án của University of Washington bởi Tianqi Chen Carlos Guestrin. Ra đời khá muộn nhưng thuật toán này thường đạt thứ hạng cao trong các cuộc thi về phân tích dữ liệu trên Kaggle. Hiện tại...
33456 sym R (7978 sym/25 pcs) 3 img 2 tbl
Kalapa’s Credit Scoring Challenge - A Naive Approach with Ranking of 19
About the Contest Information about the Kalapa’s Credit Scoring Challenge is here. After registering for the contest you can download data here. R Codes R codes for a naive approach with ranking of 19: #========================================= # A Naive Approach for Data Processing #========================================= # Clear work spa...
6403 sym R (3898 sym/1 pcs)
Pity Russia’s billionaires (complete version)
Economist-Style Plot using R Origin of the Economist plot can be found here. In searching a more simpler method of plotting, this plot can be created by only using ggplot2 as follows: R Codes for Data Cleaning and Visualization Before going ahead, you can read part 1 for creating left plot here. R codes for replicating above Economist Plot: #===...
10781 sym R (6867 sym/1 pcs) 1 img
Pity Russia’s billionaires
Economist-Style Plot using R Origin of the Economist plot can be found here. Using a very complicated approach, this plot was replicated by Le Hoang Van. In searching a more simpler method of plotting, this plot can be created by only using ggplot2 as follows: R Codes for Data Cleaning and Visualization # Data for ploting: dat <- read.csv(text...
6155 sym R (3476 sym/1 pcs) 1 img
Average annual hours worked by persons engaged, 2017
Economist-Style Plot using R R Codes for Data Cleaning and Visualization Above Economist-Style Plot can be created by using R as follows: # Load some packages: library(tidyverse) library(pwt9) library(ggthemes) # Load Penn World Table Data: data("pwt9.0") # Prepare data for ploting: some_countries <- c("France", "Vietnam", "Japan", "Singap...
5288 sym R (3198 sym/1 pcs) 1 img
Share of Vietnam government spending on health care, 2014
Economist-Style Plot using R R Codes for Data Cleaning and Visualization Above Economist-Style Plot can be created by using R as follows: #========================================== # Collect data from WHO and pre-process #========================================== # devtools::install_github("expersso/WHO") library(WHO) library(tidyverse) # ...
9580 sym R (6408 sym/1 pcs) 1 img
Vietnam ’s fertility rate
Economist-Style Plot using R The graph was created by the Economist and can be replicated by using R as follows: R Codes for Data Cleaning and Visualization rm(list = ls()) # Package for data manipulation: library(tidyverse) # Load wbstats package: library(wbstats) # A list of indicators: my_indicator <- c("SP.DYN.TFRT.IN", "SP.POP.TOTL", "...
10731 sym R (7162 sym/1 pcs) 1 img
Origin of crude oil imported to China
Economist-Style Plot using R The graph is motivated by Economist’s plot and can be replicated by using R as follows: R Codes for Data Cleaning and Visualization In this post I use uspopage data set from gcookbook package for illustrative purpose. rm(list = ls()) library(tidyverse) library(gcookbook) library(scales) my_colors <- c("#A8A9AD", "...
8017 sym R (4974 sym/1 pcs) 1 img