Trading App Analysis and Clustering
The data are provided by Revou courses. This project will try to answer these question.
- What are the user trends of our investment app? checking the trend of our apps will help check the growth.
- How much money is moving in each type of investment and how often? this make sure to apply feature in a right way.
- What kind of investment behaviour our user have in each type? investment behaviour can be different between each user and usually it wil have its own group.
- Is certain type like occupation gende ror age group have certain distinct type? this question will help in deciding our segmentation group later.
Before proceeding with the analysis i cleaned the data first. Stuff that i fix are
- change user id to str
- remove user id duplicate for safe check
- Change the datatype of date to datetime so its readable
- fill the NaN from refferal with “non Referral’ for clarity
- adding the age to age range for easier clustering in later use
For detail here are the notebook i made for data cleaning, EDA, and clustering.
Here are some the graph i produced.
Clustering
- Cluster 0 are filled with youngster and someone with less than 10 m mincome, the majority are student. most of income from non work related (undian, orangtua, tabungan etc)
- Cluster 1 are mostly filled with high income person and with age above 20>, income source gome from bisnis or payday, this correlate with occupation that is swasta and pengusaha. almost most are male
- cluster 2 are dominanly filled with woman with most of then are between 20-40, this cluster also filled with IRT occupation