Vol. 3: Embracing the Unknown – Exploring Data Science Thesis at #Otemachi

It was an honor to participate in the final Prophet thesis session. I extend my heartfelt gratitude to Mr. Kayamoto for his invaluable assistance in elucidating the functionality of the Prophet model through its implementation in Excel. Additionally, I sincerely appreciate the seven participants who actively contributed to making the session engaging and interactive.

I feel fortunate to have overcome the challenges associated with implementing the Prophet model in Excel, which deepened my understanding of its effectiveness and the interrelationships among key variables—such as trends, change points, seasonality, and the impact of holidays and events—in predicting population dynamics.

I would like to share one of the examples we worked on in Excel, where we created a curve fitting influenced by seasonality. The screenshot below illustrates the formula, showing how its effectiveness evolves as the data is updated.

In the next session, we will move on to Continuous Optimization for Structure Learning.