In one of the last weeks of the first Semester, and before going to our summer break, we had the challenge to work with the guidance of Amit Joshi and Karine Avagyan in the Digital Analytics lab.
The objective of this competitive week-long experiential exercise was to work on a prediction model for the price of an Airbnb property based on different variables. It gave us the chance to work in teams in a real-world challenge. We also had three exciting guest speakers who helped us to get motivated throughout the whole week.
First stage: data cleaning
Raw data is often messy, containing inconsistencies, missing values, or inaccuracies, so the first task was to clean it up. This involved identifying and resolving data quality issues, such as duplicate entries, inconsistent formatting, and incomplete records. We were given 24 hours and were able to see live progress between the teams. Data cleaning requires technical skills, critical thinking, and attention to detail. It sets the foundation for accurate and reliable analysis. But the most conclusion is that we had the chance to empathize with people who deal with this task daily.
Second stage: data analysis
After listening to Julie Coyette, who gave a complete introduction to how to present and visualize data in Tableau, we came back into our working rooms to leverage various analytical methods, such as regression analysis, clustering, or trend analysis, to uncover patterns, relationships, and trends hidden within the dataset to come up with a formula to predict the final price. This was the most challenging part and required a lot of teamwork to elaborate a regression that would make sense. Even climbing 1% of precision could take hours. Again, we were given a limited time and ended up with a very tight score difference between the teams.
Third stage: presentation
Finally, on Thursday, we were given the instructions for the final presentation. We had to deliver a 10-minute pitch with a five-minute Q&A with a judging panel. It was time to come up with the final solution and combine everything we had learned during the week to develop a concise and compelling logic. Expressing these data-driven insights in a non-technical but convincing way was crucial, and each team prepared key visualizations, graphs, and charts to support their findings in a really interesting way. We then gathered together to wrap up the experience. The top three teams received vouchers to use in the IMD gift shop and library.
Conclusion
This week was one of the most valuable weeks in the program. Given the times we live in, with a lot of focus on AI and data-driven decisions, it is advantageous for future leaders to harness all of this information to drive strategy.
