As Thanksgiving festivities concluded, the CHAMPS reunited to prepare for a successful System Level Design Review (SLDR). The team finished editing and submitted the SLDR report draft earlier this week and began practicing for the corresponding presentation. On Tuesday, November 28th, the CHAMPS completed their peer review session to improve the SLDR material. This class discussion provided members with vital comments to refine the overall presentation. Later this week, the CHAMPS also received feedback on the draft report and are thus, actively enhancing the class deliverables. The primary observations the CHAMPS team is addressing include a reduction of technical details and clarity regarding the problem statement, prototype solutions, and deliverables supplied next semester. Consequently, the team decided to limit the two prototype architecture slides to a general summary with explanations of terms such as long short-term memory (LSTM) networks, recurrent neural networks (RNN), and autoencoders. These techniques are the foundation of the machine learning model the CHAMPS will deliver at the end of next semester to generate warnings of anomalous events on merchant Commander site controllers.

Due to the reduction of technical details in the System Level Design Review presentation, the CHAMPS discussed the prototype development thoroughly in their weekly meeting with Verifone liaisons, Darryl Miller and Brian Russell. This meeting also allowed members to remove any sensitive information from the presentation due to the public nature of the SLDR event. The CHAMPS are looking forward to presenting on Tuesday, December 5th as they implement the final touches to the report and slides. Next week, the team will report on the event outcome and their spring semester plans.