Notes from 24th Meeting

[Joearrowsmith]

• [ ] How do we consider magnitude of movement, average trajectory velocity. Neurons correspond to velocity as well? @jonskerlj
• [ ] Which neurons correspond to x,y,z component
• [ ] Testing LSTM and linear model @linnea-evanson
• [x] Average end position
• [ ] Direction tuning, task dependent or recent history
• [ ] ICA and PCA @arahosu
• [ ] Population encoding and frequency @manfrediCastelli
• [ ] Select edge cases for our validation dataset. Trajectories that deviate away from the mean. @Joearrowsmith

[Joearrowsmith]

• [x] Are we given the task in the test dataset (We are not)
• [x] Can we use the initial position of the trial as an input for our predictions? (Not all trials start at the origin) (We are given the first data point)

[Joearrowsmith]

• [ ] Ask about the skew at the end; do we still need to predict it. Partially sorted as filtering out the last 100 ms doesnt have the end skew. But we still need to know if we need to predict it.

[Joearrowsmith]

• [ ] PCA of neurons and PCA as we go

[Joearrowsmith]

• [x] At the beginning of time is the stimulus given or not. (Stimulus is given)