AI & ML
Please note that I am still building this section. It is my hope that in time this section will grow into a set of pages for each project that will allow me to present purpose, hypotheses, outcomes, but also associated thoughts about those project in a way that can go beyond that which is included in published papers or software source code. I have yet to add content for these thoughts and works in the area of AI and ML, such as:
The comparison of human brain task / compute performance with AI performance.
What AI is missing today (why are there still things people do better)?
Building "planes" vs building "birds". When we want our AI "wings to flap" and when we don't. (I.e. engineered intelligence vs brain emulation.)
Reinforcement learning with time-difference algorithms found in prefrontal cortex vs TD in machine learning.
Machine learning for System Identification in Neural Circuits (approximating functions for neural prosthesis).
Machine learning and signal processing for the analysis of neural data from electrophysiology.
Machine learning for reconstruction in connectomics.
Design of neurmorphic processors, and application of neuromorphic integrated circuits.
Published work on probabilistic decision analysis for venture capital investments in technology companies.
Arbitrarily-precise approximation of biophysical functions using splines for fast, maximum-time-step simulations that predict the intersections of derivatives to minimize the number of scheduled events in the simulation. (Including GitHub repository of C++ and Matlab code.)
Image analysis code description (and reference to GitHub repositories).
Published work on AI assistance for an aging population.
The Knowledgetron (KT), analysis of trained neural networks and extraction of fuzzy rules.
Please check again to see content appear here about the topics above (as soon as I manage to schedule time for the necessary updates). If you already have questions for me about these topics, please feel free to contact me.