What are the most significant risks that could jeopardize the success of the project?
The most significant risk in our progress so far is the CNN not working. We have officially decided to not classify using SVMs anymore and switch to the CNN architecture, so we are investing all our efforts into this method.
How are these risks being managed? What contingency plans are ready?
Since we do not have SVM as a backup method in case CNN does not work, we would have to change our CNN method if the accuracy doesn’t reach the desired accuracy. We will be researching methods of LSTM since the input data from our new data set is varying in time length, so hopefully that will improve our accuracy.
Were any changes made to the existing design of the system (requirements, block diagram, system spec, etc)? Why was this change necessary, what costs does the change incur, and how will these costs be mitigated going forward?
We have officially decided against using SVM, so we are now only using the block diagram with the CNN architecture.
Provide an updated schedule if changes have occurred.
Eri is now working on CNN with Ryan instead of working with SVMs now.