Skip to content

gents98/Mini_Project_1015

Repository files navigation

SC1015 Mini-Project

image

  • The entire code in jupyter notebook can be found here!

Data Source

ParseHub WebScraping from:

Problem Definition

As Car Plates prices are determined by many factors such as supply & demand, uniqueness, biddings and etc, there is no solid factor that allows us to predict foreign number plates but in this mini-project, we decided to take on the challenge by reviewing how the past number plates are matched with its prices.

Motive

  • The purpose of analysing carplate prices allows buyers and sellers to have a better understanding of where their carplate stands in the market price as to avoid being short-changed in the case of buying/selling.
  • We ought to be able to provide a good estimation of a price with any given Singapore Car Plate.

Models Attempted

  • Multi-Variate Linear Regression
  • Polynomial Regression (Non-Linear Regression)

Libraries Used

  • Pandas/ Numpy/ Seaborn/ MatplotLib.pyplot/ Sklearn
  • Pandas - Groupby (Data Grouping)
  • PbPython - Natural Break (Fisher-Jenks Algorithm)

Additional Tools required to install

  • Natural Breaking ("conda install -c conda-forge jenkspy") on ANACONDA.NAVIGATOR -> CMD.exe Prompt

New Things We Learnt

  • Natural Breaking Algorithm
  • Pandas Groupby Function (Our Solution)
  • Polynomial Regression (Non-linear Regression)
  • Web Data Scraping (ParseHub Software)
  • Self-Made Features/Variables

Video

Contributions

Oh Ding Ang

  • Data Collection/ Web Scraping
  • Exploratory Data Analysis

Tan Kim Seng

  • Data Collection/ Web Scraping
  • Pandas Groupby
  • Natural Break Fisher-Jenks Algorithm

Genson Tan

  • Machine Learning
  • Data Research/ Web Scraping

References

About

Predicting Car Plate Prices

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published