I used the free time from working on the random forests to see if the ANN could be any better.
Expected output
[x] At least 20% accuracy
[x] Predict something more impactful than a random forest.
[x] ANN can do better prediction base on frames.
Actual output
[x] In reality, 3% accuracy
[x] Very difficult to predict peak stress better from the output of random forest
[x] Shows failed in terms of prediction, given that data is small per frame therefore, random forest by now is the only option in terms of a classification model.
Next Steps of ANN
Given that I have one sprint left, I will not touch the ANN until I complete a random forest.
I used the free time from working on the random forests to see if the ANN could be any better.
Expected output
Actual output
Next Steps of ANN
Given that I have one sprint left, I will not touch the ANN until I complete a random forest.