Week 10: Data Analysis and Model Debugging

This week, our team continued to make significant strides on the Verifone project as we explored and prepped the new data set we received. Ensuring that all data was in the correct format and safe for processing was a priority, and we worked closely as a team to verify these initial steps. This collective effort laid a solid foundation for further analysis and model improvements.

The Feature Engineering team took the lead on Exploratory Data Analysis (EDA) of the new data, focusing on Temporal Engineering techniques. We implemented a 20-minute interval calculation to capture evolving patterns within the data, aiming to reveal trends and outliers that could enhance model accuracy. These initial EDA steps pave the way for more advanced feature extraction, including spatial features, which we plan to explore further in the coming weeks.

Data Analysis – Events per Hour

Meanwhile, the Model Engineering team debugged the HiPerGator model to ensure it integrates smoothly with the new data. We also began analyzing distinct datasets within Jupyter notebooks to understand specific patterns from different locations. This detailed examination will guide our model adaptations and help us benchmark its performance with the new data.

On the UI side, the team focuses on refining the user interface and preparing for user testing sessions scheduled for PI (Product Increment) Day. These sessions will provide critical insights into the design’s usability, ensuring it meets user needs and project goals.

Looking ahead to next week, we plan to come together to finalize preparations for PI Day, aiming to make a strong impression with our data insights and refined model. We’ll also begin drafting the Prototype Inspection and System Level Design Review documents to outline our progress and next steps. The Feature Engineering team will dive deeper into additional EDA methods, and the Model Engineering team will benchmark our current model with the new data to evaluate its performance. The UI team will continue refining and testing the interface based on user feedback.

It’s been a rewarding week with valuable advancements. We’re excited about the progress we’re making and look forward to further accomplishments. See you next week!

Leave a Reply

Your email address will not be published. Required fields are marked *