Recently, Jack Moffat wrote up a report of his efforts to get the Brian2 Neural Network simulator up and running. He did this using Jupyter notebooks. I thought his efforts might be of use to others, and so I am linking his .ipynb file and the .html output here. I would recommend starting with the .html page as that gives an overview of what he did and what he found as the pain points. If you decide to try it yourself then download the .ipynb file so that you don’t have to retype all the code.
Brain Day is coming to the University of Waterloo this April 8, 2019. All are welcome. The event is free, and we have a marvelous line up of speakers that will cover the full gamut from philosophy of mind, to human decision making, to computational models and tools for neuroscience. Coffee and refreshments will be available from 8:30, with the talks starting at 9:00 AM. The venue will be in the E7 3343. Look for E7 on the campus interative map. A poster with the schedule for the whole day can be found here.
- Breaking Down Attention: Distinct Contributions from Neurons in Different Brain Regions
- Procedural Creativity in Art and Science
- Computational Challenges of evolving the language-ready brain
- Unifying models and theories of human brain function and dysfunction
Hope to see you there.
The group has just had an article published in Frontiers on Statistical Learning Impairments as a Consequence of Stroke. The original goal of the research was to assess whether people with right hemisphere strokes might have more trouble updating a statistical model of language than participants with left hemisphere damage. Along the way things got more interesting and murkier. We ended up running many variants and including older and younger controls. So there are a lot of data, and they are available for you to look at too.
In summary, the main findings are:
- Young adults have a capacity limit for learning statistically defined languages.
- This capacity limit was attenuated for older controls.
- Participants with stroke, left or right, usually had significant deficits learning this material despite intact language comprehension, functioning well in their home environments, and being a relatively long time from the stroke itself.
- No clear location was associated with the magnitude of the deficit, but the strongest (though non-significant) association was near the anterior insula/superior temporal gyrus.
In addition, the work benefited from the work of an undergraduate (at the time), and a former PhD student so it was a pleasant spanning of academic generations, and utilized materials provided by Dr. Richard Aslin so it was also a tribute to collegiality and collaboration.
Please feel free to send on any of your comments or critiques.
View older posts in the Archive.
subscribe via RSS