How I Spent My Day | 08-05-2025

It was indeed a hectic day for me because all my assessments were due already and needed to be submitted; also, I had a client that I needed to work on her task as well. I had to run an analysis on crime hotspot detection using a machine learning algorithm and also integrate a map into it for visualisation since I was using the longitude and latitude as coordinates.

Training_Time_Comparison.png

So the task was a heavy one for me, and I still had to write a full academic report which included the abstract of the research, the introduction, literature review, methodology, analysis and discussion, result presentation and then conclusion, recommendation and future direction of the studies. It was an interesting work, and I am glad I did it till the end.

My intention is to learn how several machine learning models work, such as the XGBoost, Random Forest, Support Vector Machine, Logistic Regression, Decision Tree and even the K-Means model. Most of the analysis I did was a comparative analysis to determine which of the algorithms performed better.

In one of the analyses that I ran, I noticed that the support vector machine took a lot of time for training my dataset, and it was a big limitation to my research, and since I didn't have the whole day, I decided to drop the model and look for others that are much faster in carrying out the task. I was able to replace it with logistic regression and one other machine learning model.

It was both tough and a beautiful experience, if you ask me, because I was able to learn a whole lot during the process. I intend to write and have this paper published, and that is why I'm working tirelessly to ensure that I get everything right. So that was how I spent my entire day on my system trying to get my work and that of my client ready for submission.

So, guys, I will see you all in my next publication. I hope you enjoyed my story of how I spent my day. I will always update us all on how my entire day went. Until then, please stay safe and continue to Steem on.

Sort:  

Upvoted! Thank you for supporting witness @jswit.