Today there is a new computing frontier – business analytics.
Some people may be interested in the relationship between our brains and the new Artificial Intelligence that is impacting the business and research worlds.
In an earlier time we used computers to help us perform computations, then to organize data, later to visualize data so that people could better see correlations and make decisions.
Now we are moving into the next generation. Here, analytic programs such as Watson by IBM (IBM), Microsoft (MSFT) Azure Machine Learning, and AWS Elastic MapReduce service by Amazon (AMZN), are sifting through enormous amounts of big data to answer business and scientific questions.
According to Wikipedia:
Analytics is the discovery and communication of meaningful patterns in data.
In the end, all these are sifting through the data trying to abstract out significant relationships between people and/or events.
As I noted in another post, Abstraction is what the brain does.
Some people may be interested in the relationship between our brains and the new Artificial Intelligence that is impacting the business and research worlds. In this post I will describe, very briefly, how brain cells work together to create knowledge.
Neurons – the basic elements of the brain
What follows is a gross simplification of the systems, but it will give the reader a general idea of the concepts.
The brain is an enormous collection of interconnected neurons. Neutrons are the biological cells that transmit electro-chemical signals throughout the body and collectively, in the brain, coordinate movements, create memories, emotions and intelligence.
The neuron is different form all other cells in the body, most obviously in their from. It consists typically (there are exceptions to almost everything here) of a body with a very long appendage called an axon. It is along this axon that signals are sent. Like any cell, the cell body includes a cell nucleus.
Also, attached to the body, there are a set of structures called dendrites, like a bunch of spiky hair, branching out every which way from the cell body. These are responsible for receiving signals from other neurons. One neuron may receive input from a sensory organ (touch sensor for example), from just one other neuron, or from hundreds of neurons.
Similarly, the end of the axon branches into multiple branches each ending in an axon terminal. Again, an axon may terminate on an effector (muscle or internal organ), on just one other neuron, or on many. The purpose of the terminals is to either stimulate or inhibit its target.
When a neuron is stimulated sufficiently to reach a threshold, then it will initiate an action potential, an electrical pulse that travels down the axon to all the terminals. (The electro-chemical process by which this is done is one of the most extraordinary and beautiful constructs of nature.)
Each terminal connects to its target at a synapse, where there is actually a synaptic gap between the terminal and the target dendrite or muscle, etc. Communication occurs when an action potential reaches an axon terminal. This triggers the release of one or more neural transmitters into the synaptic gap, where they drift to receptors on the dendrite terminals of the target cell. These transmitters are special chemicals that will either excite or inhibit the target cell.The brain consists of a vast network of interconnected neurons, many with feedback loops. It is estimated that there some 85-100 billion neurons and up to 100 trillion synapses. That is a lot of connections!
The scientific obsession with neurons really began at the end of the 19th century. Spanish anatomy professor Santiago Ramon y Cajal used a special dye to stain brain tissue. Under the microscope, neurons were revealed in exquisite detail. “A dense forest,” Ramon y Cajal called it a field of little branching cells that would soon be named neurons. [GPB]
How it all works
So, whenever some particular set of stimuli come into the brain from the senses, then a particular set of neurons is stimulated. Let’s say you are looking at an old fashioned water glass.
- This will form a unique pattern of activity among the neurons in your visual cortex – a unique network of firing neurons
One thing about the neurons is that (at least in some areas of the brain) whenever one neuron interacts with another, the interconnection is strengthened – any repeated firing will stimulate or inhibit just a bit stronger than the last time.
- Thus, the same stimuli, if repeated, will tend more and more to activate the same network of neurons – and this pattern or network of activity will come to represent the object or event.
So, if the same stimuli (say the image of your glass) is repeated, or perhaps the same sequence of events (learning the 5 times table), then that particular pattern will be enhanced and more easily replicated. The same pattern will be induced by a weaker or non-exact stimulus. You recognize your partner from a silhouette. You no longer need to see a fully lit face.
Also, this pattern will be very similar to that produced by other glasses so the object will be recognized. The reason for this is that, over the years, hundreds of images of glasses have set up a network pattern and this glass, due to similarities, has set of the “generic drinking glass” pattern.
If you are at a party and have the only green glass, then further firings will link up to the concept of “mine,” and you will know that it is indeed your Black Butte Porter you are about to drink.
All in the patterns
The key to all this is the patterns of firing.
- It is the strengthening of connections that firms up a given pattern – and this is what learning is.
These patterns are not static, but change over time, and the whole dynamic progression plays out, recalling, perhaps, your favorite scene from The Twilight Zone. Again – it is the dynamic pattern over time that forms the memory.
These are the keys to memory and learning. They can be interconnected. If your grandmother had a set of the same green glasses, encountering this one glass might stir up thoughts of home made blintzes on a winter day.
They can also influence behavior, as with Pavlov’s famous pooch. Here the sound of a bell became associated with food, and it alone would be sufficient to start the dog salivating.
Mining hidden data
So if we return to the tree planters problem, how to differentiate the White Pine from the Jeffery Pine without counting the needles, then we can begin to get a picture.
The brain creates a complex pattern of activity when seeing the natural seedling that needs to be identified. This pattern has subtle differences according to many features of the tree. Below the conscious level, the brain begins to weigh each of the many factors until a preference for one pops out and a message is sent to the conscious brain “Western White Pine!” Still, however, the conscious mind has not idea how this conclusion was reach as there seemed to be no analytical process.
Back to modern analyticsBut actually there had been an analytical process. Down at the activity pattern level, the little neurons had cooperatively been computing a statistical analysis, weighing multiple factors to calculate the answer.It is precisely this type of behavior that modern analytics is trying to replicate.Let’s go back to the Wikipedia definition above:
- Analytics is the discovery and communication of meaningful patterns in data.
The current systems attempt to find patterns that are hidden in gargantuan quantities of data – much more than any one human could ever hope to digest. This is what the wondrous brain does. Finds patterns that help us survive.
This is what we are trying to do with our new creations.
— Comments Appreciated —
Elegant, Handcrafted, Genuine Leather
Soon for iPhone 6 Plus!