Summary
This project involved working in a group of three to analyse various LEGO datasets. Our goal was to create a price prediction model that could predict the resale value over time of a given Lego set based on various factors (like set size, mini-figure count, theme, etc). We also used a full end-to-end pipeline to create this project, our data was stored in AWS, analysed and then some aspects of it were presented in a Dashboard. We used version control (Github) to store our data and code. Multiple different machine learning models were used to create the price prediction such as linear regression, random forest, clustering and XGBoost, with Random forest performing the best overall.
My personal goal with this project was to really stretch my understanding of the end-to-end process nature of data science projects, preparing me to contribute effectively to larger, real-world projects in the future.
Key Skills
-
Python
-
Knowledge Embedding
-
Sklearn
-
AWS
-
API's
-
Basic Dashboard (with Dash)
-
​Cleaning, analysing and visualising data from multiple datasets​
-
Team management
