Pet project 👾

Side win prediction based on score, duration and heroes data

The idea of predicting the outcome of a match after its completion may not seem very interesting and applied.

But I chose such a task in order to go through the main stages such as:

• data scraping
• data analysis and cleaning
• building several models and choosing the best one.
• hypothesis testing

I presented the implementation in the streamlit cloud - link

The detailed steps are described in the notebook and .py files

---

Snippets from streamlit