START HERE: This website is about demonstrating how to properly capture a periodic waveform. All too often we rush to use the Fourier Transform to look for a 'frequency response' in a waveform behavior. We are taught to consider 'infinite limits' and 'infinite sums of points'. The 'after thought' is to focus more on the practical truth: that the best we can do is measure a behavior, over a 'finite time limit',and with only a 'finite, discrete, and bounded' set of points to work with, at the end of the measure. And, this website will show, that taking time to actually consider the finite on its own terms is enriching, rewarding, and for the most part, very real. The examples given in this website range from a mixed set of integers to sounds from insects to the Electrocardiogram to the behavior of network traffic in communications. And, it is the word 'communications' that we have leveraged to give us the ability to capture that periodic waveform, and the transition into the next periodic waveform, and so on. The 'trick' is to consider the waveform as a communication and then learn how to bend that communication into its periodic pieces. And, once that is achieved, then take a better informed discrete Fourier Transform, because it is now isolated into periodic pieces as the Fourier Transform is designed to work with. Or not. After a while, one may just learn to ride the wave of periodic transitions. This website is not about discouraging the use of the Fourier Transform, on the contrary, we will show we need to cascade 'Two Fourier Transforms' in a row, which may loosely be called 'The Cepstrum', then step back, and take a look.
There's a little more in our first folder, An Introduction to Data Plasticity and the Bending of Number, essentially, stating some background for, and thanking a couple old guys, Jim Cupples and Randall King, who, motivated the 10 or so chapters on this site. Go here for a PPT on Tukey's EDA and Bloom's Taxonomy. You can even listen to what Kristann Loken (Terminator 3) was communicating to the Data Warehouse at 300 baud/sec, and then go to the next slide to hear how she sounds in a modern communications room.
Our second folder instructs how to carry out The Bending of Pseudo Random Data using a simple example of only 243 numbers (see below). In part, a Dr. Misty Blowers motivated this with her 'Random 50 test'. It ends with a teaser, to the remaining topics; we bend the sound of a Wasp.
The fourth folder is called the Cepstral Bend on Fourier. The first part starts out with the Fourier Transform and takes the time to emphasize exactly why 'Period' and Fourier Transform' go together, using numerous simple plots, and it emphasizes when to expect error, which is when there is a waveform that was chopped off before or after the periodic, not on it. The second part gives an ad-hoc definition for Cepstrum by example. It shows the secret to finding a 'best' measure of periodicity. The word ad-hoc is inserted because there are technical definitions given to explain the reasoning behind the usefulness of the Mathematical definition of Cepstrum. I prefer my Ad-hoc definition in practice. And, I prefer my explanation as it gets to the root of why this thing works in the first place. There's room for improvement.
The fifth folder folder simply gives more examples of The Reframing of Quantized Bleets (of Sheep) and Buzzings (of Bees, Mosquitoes, and the cricketing of the Cricket). I left out the BAT.
The sixth folder, Reframing EKGs through Quantized QRS's shows an Abstract and Slides to a presentation given in October of 2016 at the Applied Imagery Pattern Recognition Workshop, at the Cosmos Club in Washington, DC. It will demonstrate how we look at the electrocardiogram in the light of Cepstrum, which has been done before, but probably not to this degree.
The seventh folder, the Dynamic Plasticity of an SVD (Singular Value Decomposition) departs a little from the main focus of waveform, and shows the plasticity of a 'nothing space' (null-space) waveform with an example using matrices. It gives some papers, including one on independent components and another on Finite Field Theory, that in no small way have influenced the making of this site.
The eighth folder, The Compression of multi-layer Neural Networks with JPAT. JPAT is device that nay convert a conventional NN into a Bidirectional NN, and it may compress it to satisfy the Universal Approximation Theorem (I get the word 'may' from Ms. Carol Ann, an editor of the Patent Application), and it maybe used to connect several NNs internally.
The ninth folder is the first time we look at "Computer Speak". The folder is called Physical Sensing of Computer Network Traffic with AVIPE. AVIPE is a way to see and hear the idiolect of moving 1's and 0's. Yes, there is a way to BEND a streaming Computer Language.
The tenth folder is probably the culmination of everything presented, The RueBea System for HCI. It
is just being put together. The goal; the idea; the end state: A Human
Computer Interface that is built upon an interchange of Human and
Computer idiolect based on the bending and association of waveform, in
discrete, finite, bounded, Cepstrum-based language. It's a start. The big picture idea is found in the DARPA Innovation House Poster below.
The eleventh folder is called Lagniappe, and
among the items is a basis for five Lesson Plans that have been, taught
at the Elementary, Middle, and High School levels. (Primarily at the Louise S McGehee and Jesuit High School in New Orleans, but began at St Amant of Gonzales, LA.
The twelfth folder contains Resumes, outlining my experience with Signal Surveying and Diagnostics. One with a Biomedical lean, and another one has a Signal Intelligence lean. Also, included is a strategic mission statement, written for me by another old guy, George Ackmon; he saw what I was doing instantly.
