Resale-prices-in-singapore

• Resale-prices-in-singapore
  • Description
  • Takeaways
  • Areas to improve on
  • Heroku deployment
  • Support me

Description

V1

This project is my first attempt at implementing machine learning and analysis techniques I have learnt. At the time of creating this project I had just started my maching learning course in my school. The goal of this project is to create a model to predict the resale prices in Singapore. I believe that the best way to learn is to venture out of comfort zones therefore I attempted a raw public dataset.

V2

Fast forward a few months and I'm looking to further enhance this simple dataset. The next focus would be proper code documentation and deployment. I will be deploying this as a streamlit, coupled with CI/CD techniques.

Takeaways

While building this, I have improved my data wrangling, cleaning and pre-processing which I might not have learnt building the models in class. I've also learnt hyper-parameter tuning and XGB.

I have referred to countless articles and other github users for referrence to build my model.

I was motivated on the lat lon model when I read an article on towardsdatascience regarding openmaps api and they too had a similar project.

Areas to improve on

Improve foundations on data manipulation and wrangling. Struggled to make the right outputs and dtypes.

Improve code and understand parameters.

Learn more models and understand the capabilities of the more commonly used ones.

As it is my first public model, I am still currently still learning the PEP8 style.

Heroku deployment

Try it out here! Might take awhile to load if it's inactive

Support me

Please do leave feedback and criticism.