Duke Institute for Brain Sciences Methods Meetings
Workshops and tutorials on methods, statistics, and models in neuroscience.
Category programming
Experimentalists often find themselves needing to present carefully controlled stimuli to participants, control and catalog stimulus conditions, responses, reaction times, and other empirically variables of interest. While tools such as...
This tutorial introduces the computational cluster at the Duke Brain Imaging and Analysis Center (BIAC) and presents a few simple use cases. Hopefully it will make your experience handling/analyzing neural...
In this tutorial we’re going to talk about what probabilistic programming is and how we can use it for statistical modeling. If you aren’t familiar at all with Bayesian stats,...
Today we’re going to tackle a common problem faced by graduate students: you brainstormed to design an experiment with your PI, toiled away programming the task, put it out into...
As human beings continue to march through the information age, we produce ever-larger quantities of data every second. This is especially relevant to us as scientists, as data gathering and...
Hello! I hope you’re doing well. Today we have a tutorial on ~ neural networks ~
A too-brief overview of this powerful and (usually) intuitive approach
to statistical computing.
There are lots of powerful things you can do with the Markdown editor. If you’ve gotten pretty comfortable with writing in Markdown, then you may enjoy some more advanced tips...
Why git? git is a simple but highly flexible system for keeping track of stuff on one or more computers. But, everyone probably already has some way of doing the...
Category professional development
with Jekyll and GitHub pages!
If you’re making a blog post on R-related content, you’re probably going to do it using R markdown (Rmd). However, you may have noticed that our website runs on GitHub...
There are lots of powerful things you can do with the Markdown editor. If you’ve gotten pretty comfortable with writing in Markdown, then you may enjoy some more advanced tips...
Category statistics
You’ve probably come across linear regression from time to time in your research, or in reading papers – but how does it work? What is linear regression? What are the...
If you have ever tried to analyze time series data, you know that time series present all kinds of statistical challenges. Probably the most challenging aspect of time series data...
Signal detection in math and psychology
So, you’re designing an experiment and you’re faced with answering the age-old question: How many participants do I need for this experiment to work? Probably, your advisor sent you down...
As human beings continue to march through the information age, we produce ever-larger quantities of data every second. This is especially relevant to us as scientists, as data gathering and...
Have you ever ran a regression and wondered where the coefficients come from or what they mean? Or perhaps you’ve tried the same analysis with different coding schemes, and the...
Bayesian statistics are gaining a whole lot of traction in psychology, neuroscience, and a whole lot of other fields. But, since most psychology departments don’t teach Bayesian statistics, you probably...
A summary and some questions for group discussion.
Category journal club
A summary and some questions for group discussion.
Category machine learning
Topics/Organization: Some geometric intuition for SVMs Introducing slack variables (Soft-Margin SVMs) The SVM loss function (primal and dual forms) SVMs and Kernels Some coding examples of the above (for fun!)...
Hello! I hope you’re doing well. Today we have a tutorial on ~ neural networks ~
Category math
The beauty of the Fourier series and Fourier transform
Finding structure The whole goal of our experiments is to uncover some structure in behavior. In some cases we can make this easier for ourselves by simplifying the data we...
If you have ever tried to analyze time series data, you know that time series present all kinds of statistical challenges. Probably the most challenging aspect of time series data...
If you’ve worked with any kind of neural data, convolutions were involved at some point. Having an intuition for convolutions is quite useful in thinking about what your data actually...
If you’ve ever wondered how R or python gives you regression coefficients, the answer is linear algebra! Linear algebra operations are essential to almost all modern methods for analyzing or...
Category psychology
Category neuroscience
If you’ve worked with any kind of neural data, convolutions were involved at some point. Having an intuition for convolutions is quite useful in thinking about what your data actually...
Category modeling
Topics/Organization: Some geometric intuition for SVMs Introducing slack variables (Soft-Margin SVMs) The SVM loss function (primal and dual forms) SVMs and Kernels Some coding examples of the above (for fun!)...
Finding structure The whole goal of our experiments is to uncover some structure in behavior. In some cases we can make this easier for ourselves by simplifying the data we...
If you have ever tried to analyze time series data, you know that time series present all kinds of statistical challenges. Probably the most challenging aspect of time series data...
Category tutorial
Plotting in Python: A quick rundown
The curse (or challenges) of dimensionality (p>>n)
Two methods meetings ago, we learned all about linear regression. Probably the most common statistical technique, linear regression is most often used for predicting a continuous dependent variable (e.g., “What...
Analyzing data with repeated observations for a particular participant, stimulus, or other group is one of the most common things you need to do in psychology & neuroscience, like most...
Today I am going to present on an alternative way to analyze Likert scale data by using ordinal regression instead of linear regression. But first, why is it even a...
Last week, Raphael presented a fantastic conceptual introduction to drift diffusion models, which are an extension of signal detection models over time. Here I’ll be talking about what model fitting...
Category signal processing
The beauty of the Fourier series and Fourier transform
