The project is divided into 2 parts, which are mainly the following:

Part 1

​

Data - This section will provide a description on the raw data we have extracted, the various processes employed to clean the data, the rationale on categorising and dropping certain data, and the codes the were executed to derive the additional fields. 

​

Part 2 - For the Analysis sub-sections, the data, tools, analysis and inferences will be discussed. 

 

Descriptive Analysis - Using python coding, descriptive analysis was performed on the cleaned data to show general trends in the real estate market.

 

Dynamic Dashboard - Leasehold properties are at risk of lease decay! To address this belief, a dashboard was created to chart the performance of properties that we launched in 2008, 2013 and 2018. Users can use the dashboard to navigate the trends of these properties to draw their conclusions.

 

Inferential Analysis - Yes, it was observed in section 1 that the Profit/Loss (%) is higher for the first buyer (transaction 2) as compared to subsequent buyers (transaction 3 and beyond). By deploying inferential analysis, we can ascertain whether these observations are significant, and from the various permutation of variables to derive further insights into the property trends across districts, floor and tenure.

​

Condo Price Predictor; Machine Learning - The dataset was used to train the prediction machine and different algorithms were deployed. An assessment was made to determine which machine learning model was the best-fit. The best-fitting model was then transformed into an app for users to make prediction on the transacted price based on their selected property, size, postal code, and year (2024-2027) & month.