Week 9

During last week’s meeting on Friday, Karsten gave me a lot of pointers on how to approach the problem of building a classifier from a discrete signal (like those of the Muse headband). The discussed techniques were smoothing, using (sliding) windows, …

The days after that meeting, I tried to link my previously constructed SQLite database to the extracted discrete brainwave signals instead of just writing these to plain text files on external memory. Surprisingly enough, the performance of the database approach proved to be far worse than the plain text file alternative, due to the fact that the corresponding write buffers have to be reopened upon each insertion in the former case.

After having disposed of the database idea, I motivated myself to start implementing the techniques that me and Karsten talked about in the meeting. First, I smoothed the brainwave signals to regular intervals of 1 second. Afterwards, I imported an Apache library to help me calculate some descriptive statistics of a discrete signal. In time, these statistics could then be used to build a Weka classifier, which brings me to today…

Currently, I am working on a concrete (visual as well as computational) interface to keep the training and the classifying phase clearly separated. Finally, the Weka classifier itself should be constructed with the appropriate training data, after which it can be used to classify some previously unseen signal input in a basic prototype. Although this sounds to be a fairly reasonable deliverable by Monday next week (the day of my presentation), I expect there to be some more unforeseen challenges along the way.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s