How I Spent My Day | 26-06-2025
Hello guys,
I welcome you all to my blog. It is another beautiful day of the week, and I'm delighted to share with you how I spent my day. A diary is a way to help me keep track of my daily activities and also help me keep track of my progress.
Yesterday, I was supposed to be at the lab for a meeting with Professor Celesine regarding one of the major projects the centre is handling. However, after carefully considering the limited time I have to ensure that my assessment is finished, I decided to stay back at home and work on the task.
So when I woke up that very morning, I did all I could do, such as tidying the entire place, preparing breakfast, and a lot more. Immediately, I was done with all of that. I took my bath, and then I sat on my system and started working on it.
I was on my system for a very long time because I was troubleshooting the entire system to handle the errors coming up with the models. The first thing I noted was the limited data fetched by the OpenWeatherMap API, which wasn't good enough to train machine learning models.
That made me change the dataset used in the analysis because, aside from the real-time data from the API, I also added secondary data from Kaggle, which made the data reasonable enough to be trained for machine learning and the hybrid ARIMA-ML model.
In my findings after the analysis, I noticed that ARIMA, which is a statistical model, performed well only on linear relationships but performed very poorly in complex data or non-linear relationships, where we may have lots of weather data combined for measurement.
The ML and the hybrid model performed better in the non-linear relationship, but the hybrid was the best. So at the end, I was able to write the chapter that involved analysis and discussion. I'm still on it because I have a few charts that need a clearer explanation.
After a long time working on the system, I decided to rest my head a little because it has been a stressful day for me. So that was how I spent my day. See you all in my next publication.
🎉 Congratulations!
Your post has been manually upvoted by the SteemX Team! 🚀
SteemX is a modern, user-friendly and powerful platform built for the Steem ecosystem.
🔗 Visit us: www.steemx.org
✅ Support our work — Vote for our witness: bountyking5
Hey @simonnwigwe, thanks for sharing a glimpse into your day! It's fascinating to see the behind-the-scenes work that goes into your projects, especially the challenges you face and how you overcome them.
That deep dive into troubleshooting the ARIMA-ML model and finding solutions with the data from Kaggle and OpenWeatherMap API is really insightful. The comparison between the ARIMA model and the hybrid model is especially valuable. It's a great learning point for anyone interested in data analysis and machine learning.
Keep up the excellent work! I'm looking forward to seeing the final chapter with those clearer chart explanations. I'm sure it will be great.
For other STEM enthusiasts on Steemit, what are your experiences with ARIMA and hybrid models? What tools do you use for troubleshooting? Let's learn from each other!