I’m not a Data Scientist
First things first, you know, I’m not a Data Scientist, even thought, I have a strong opinion that “Data Scientist” is just a fancy term for a software developer who knows how to dig into some large data sets.
Either I’m a Artificial Intelligence scientist, the ones you can image, wearing white suits, and working at a lab, exploring the human brain, just like you can see at the “Transcendence” movie.
I’m just a developer, who has some years of studying Human Psychology, patterns behaviors, statistics and so on.
Now, with that in mind, what you are gonna read inside this article is just a bunch of “raw thoughts” about Artificial Intelligence, what I think about the Human Mind it self, and what tools we have today for doing some nice stuff, like sentiment analysis, and so on.
##
Your brain is a CPU
Ok, so, how does the Human Mind works? The brain it self is just a bunch of neurons, biological things, that transmits low voltage electricity (not the one you have at your house, neither the one you have at your car) from one neuron to another, and with that, makes all the processing unit, you could think about it just like a huge CPU, yeah, your brain is the most advanced Central Processing Unit, well, at least at 2017, I know we have all kind of data centers, with huge computers, super computers, and so on, but hey man, this piece of the human body controls everything inside your body, you may argue that the heart can beat it self without the need of the brain, but what happens when it needs to increase speed? What happens when you need to talk? You can’t tell me that your vocal cords just move it alone, they need something controlling the flow of air that goes out from your lungs.
How we tell someone is conscious?
Human communication, this is the first thing we could use to tell that someone is conscious of the world around him/her. The answers to the environment stimulus tell us that there’s something inside this body that can feel a stimulus, process it in a way it understands, think on a stimulus answer, and stimulate the world around with the answer, either it would be voice, either a raising hand, either running.
Let’s go back to the baby age, and baby language.
When you are a baby, people don’t understand you, even thought, I’m pretty sure babies have their own language set, that they use to communicate thought each other, which again, you can find some nice jokes about it at some movies, and series. But we have the need of the baby learning our language, so we can communicate with them without the need of signaling things, in the hope they would understand, that’s where my “Basic Human mind work mechanism” theory starts.
##
IN and OUT Stimulus
As with computers, we humans have INPUT and OUTPUT “peripherals”, INPUT sensors, and OUTPUT “peripherals”. We could imagine that yours eyes, yours ears, yours skin, and so on, they are all INPUT sensors, and that yours hands, yours foot, yours vocal cords, they are OUTPUT peripheral. Keep that in mind, as I’m gonna use these a lot in the examples above.
“Basic Human mind work mechanism” theory
So, everything is a stimulus, either it is visual, or audio, or tactile, any one of them, it’s all stimulus, your mind learns that some stimulus is something, for example, when you start learning a language, either it would your mother language, either it would be a foreign language, you start connecting images, with words, for example, your mother points to a tree, and says “tree”, your mind connects the image of the tree, the visual stimulus to your eyes, with the word “tree”, which is also a stimulus, now, a audio one, later, when you start learning how to write, you connects again, the audio stimulus “tree”, with a series of “moving your hand” out to the world stimulus. I know, I’m using the word “stimulus” a lot, but that’s all about Humans, Humans are all about stimulus.
Back to the theory it self, this series of converting INPUT signals to a data format your brain can parse, and classifying then into categories is what makes the human mind do every kind of things.
For example, bad experiences, may be categorized at a file named “NOT TO DO” inside your mind. Language things, would be stored inside another file named “LANGUAGE STRUCTURE”, and so on.
Now, I came to you guys, with a small question, how does the human brain, takes all this raw data, and categorizes it inside so many places? And take decisions with this?
Let’s imagine the “bad experience” thing, you, as child, got your finger in the wall socket, you got some electric shock, we can now imagine the mind’s algorithm categorizing a lot of data, for example, the image of the wall socket, not only the image it self, but the socket pattern, common things that may be near it, also placing, and so on, attaching it to the feeling of “pain” that the electric shock gave you when you connected your finger with this object, all that raw data got categorized at the “NOT TO DO” inside your mind, now, the next time you sees a wall socket, your mind instantly runs another algorithm, this algorithm recognizes the pattern of the wall socket, and selects everything that looks like this pattern finding that inside the “NOT TO DO” file, you at a past time did a action that led you to have a bad experience, this trigger some other algorithm, the “decision make” one, telling you not to put your finger, or any other part of your “INPUT sensors” inside this object. That’s basically how every decision takes place inside your mind, a series of algorithms with patterns recognition, INPUT sensors data analysis, connecting all these RAW data to make you do or not do something.
Everything is about taking a older pattern, categorizing it, connecting with other experiences, and taking decisions on it. That’s how human minds works, and that’s what we try to make algorithms looks like.
> Source: http://devhumor.com/media/we-added-ai
“What about things that aren’t categorized” theory
Ok, you got this far at the article, that’s nice to hear, thanks for trying to understand the chaotic thinking of this human mind. Now you start to wonder, and what about the things that aren’t categorized? New experiences where you need to take a decision about? How does the Human Mind take a decision on it if it doesn’t have a recognized pattern, or so on? Simple, it doesn’t take a decision, that’s where we come with our doubts, when your mind can’t recognize any pattern, anything that got from the past to apply to this situation, and I mean, anything at all, after all, we are talking about a huge CPU which can find patterns where you can’t even imagine, mixing categorized data, with some RAW data to index this and categorize even more, when your mind can’t recognize anything, it just fails, simple as that, it just start doubting, and that’s where some other “stimulus” come to rescue, the “self-preservation stimulus”, the level of courage you have to do unknown things, a friend’s suggestion of “go on”, or “don’t go”, the algorithm now takes these as the final and only decision maker, and reacts in some way.
For example, you are being robbed, the robber has a gun, your brain have never been in a situation with a gun before, it doesn’t know how to react properly, not from its own data, but it does recognizes the “gun pattern” from a picture you saw in a book, then it connects it with a warning you read from a poster “guns kill”, now, what should you do? what decision should your mind take? That’s where we get the “self-preservation stimulus”, your mind takes a decision based on keeping alive, after it recognized the gun pattern, and connected it with the single thing it has available, that guns can kill, based on that, it can use some other decision making items, for example, the level of courage you have, based on this, your mind can tell you to freeze, and give the robber everything you have, or to react, with a “Hadooken” and run from the assaulter.
You see, everything inside the human mind is about getting inputs, categorizing it with know patterns, that can be from a variety of sources, and making a decision of the output it must take.
Which reminds me of something that we know, yes, computers and computer software.
The only difference is that the mind software is running over a high level CPU, and can store petabytes of raw and categorized information, running pattern recognition over all this data. That’s where human kind is failing to run a real A.I., we don’t have enough CPU available to embed inside a small machine, we can surely run some AI inside a huge Supercomputer, but hey, I don’t have a supercomputer at home.
Babies, Children, and Adults, growing your data set of patterns
<TO DISCOURSE HERE>
Sentiments: Analysis
Now that you know how the human mind works, you are most than capable of doing your own AI code, always remember, takes a input, discover the patterns over it, if found, make a decision, if not, find something simular, runs some last decision makers, and do a output.
But what about Sentiments? How do we classify them? Again, same theory, except that now you have some other patterns to classify. In today’s world sentiment analysis is the act of getting some data, and
systematically identify, extract, quantify, and study affective states and subjective information
Source: https://en.wikipedia.org/wiki/Sentiment_analysis
That data can be a sound, or either a text, your software names it.
<<TO CONTINUE HERE>>
Learning Deeply
> Source: http://devhumor.com/media/deep-learning-is-fun
Deep learning (also known as deep structured learning, hierarchical learning or deep machine learning) is a branch of machine learning based on a set of algorithms that attempt to model high level abstractions in data. In a simple case, there might be two sets of neurons: ones that receive an input signal and ones that send an output signal. When the input layer receives an input it passes on a modified version of the input to the next layer. In a deep network, there are many layers between the input and output (and the layers are not made of neurons but it can help to think of it that way), allowing the algorithm to use multiple processing layers, composed of multiple linear and non-linear transformations.