This note is just a collection of past useful notes to know to apply machine learning methods for the analysis of topics interesting in the neural sciences.
Estimators
You need to know all Parametric Modeling. We want to estimate unknown random variables with some observations.
Maximum Likelihood
See Bayesian Linear Regression.
Bias-Variance Decomposition
Fisher Information
See Parametric Modeling#Fisher information.
Applications
Echo-locating bats
Bats use echo location to locate the target, a platform. Bats use to click slightly left and right compared to their direction of motion.
Bats Locking behaviour
Before looking the bat uses right and left click, probably to look for the target (you can have the emission intensity). After the looking it becomes a single distribution.
Tuning Curve Neural Decoding
The fisher information reaches a minimum at the peak of the curve, that is where the neuron likes to fire. The neuron is more informative in regions where the tuning curve is steep. This is an argument against the tuning hypothesis, where the idea of neurons acting as feature detectors that fire when you expose them to preferred stimulus, or most stimulus-specific information about sound frequency when they fire at that frequency.