In this article we are going to understand about Artificial Intelligence basic hierarchy based on the context of a human, because humans are the most intelligent living being stands on the earth. The main goal of Artificial Intelligent is to build systems that can function intelligently and Independently. AI can be recognized as a main sub domain of computer science.
A program that can sense, reason act, and adapt can be defines as Artificial intelligence. Making computers do things which require intelligence.
Human can Speak and Listen which enables communication with each other with their familiar languages. This is the field of Speech Recognition, much of speech recognition are statistical based. So, this speech recognition is a field of Statistical Learning.
Human can Read and Write text with their familiar languages. This is called as NLP which stands for Natural Language Processing which also belongs to Statistical learning. Statistical learning belongs to Machine learning which will be discuss later in this article.
Human can See with their eyes and process what they see. This is a part of Computer Vision. Computer vision falls under symbolic way for computer to process information. Which is called as Symbolic Learning. Humans recognize different types of scenes around them which creates images of that Environment. For example, when a human saw a garden it recognizes all the components of the garden like Trees, Flowers, Insects, etc. So, this is the field of Image Processing which is not even related to Artificial Intelligence. But Image Processing is required to computer Vision
Humans can understand their environment and Move based on the environment. For example, Walk, Run, Sit etc. This is the field of Robotics. Which belongs to Symbolic Learning.
Humans have ability to Understand Patterns which is belongs to the field of Pattern Recognition. For example, grouping of objects like separating two different shapes rectangles and triangles.
Machines are better in pattern recognition compared humans. Because machines can use more data and dimensions of data. This is the field which belongs to Machine Learning.
Let us see about human brain. Human brain is a network of neurons and human use this neuron network to learn new things. If we can recreate the structure and the functions of the human brain, we might able to get intelligence capabilities in machines. This is the field which defined as Neural Networks in Artificial Intelligence.
When these neural networks are more complex and deeper, and when user apply those to learn and observe things. That is the field of Deep Learning. Which is categorized under Machine Learning. There are different types of deep learning and machine techniques which are each different from one to another to recreate what the human brain does through a computer-based system.
If we get a network to scan images from Left to Right and Top to Bottom. It is called as a Convolution Neural Network (CNN). Which was categorized under Deep Learning. This is used to recognize objects in a scene. Where Computer Vision will be used what we have discussed before. So, then Object Recognition will be applied through Artificial Intelligence.
Human brain can remember the past for example “I have left the School in 2015”. Like same as human brain we can make a neural network to remember past. These neural networks are called as Recurrent Neural Network which is a subdomain of a Deep Learning.
Now we can understand that there are two ways Artificial Intelligence can be used. One is symbolic based, and another is data based for the symbolic based side we use symbolic learning and for the data-based we use machine learning.
In machine learning We need to input lot of data so it can learn. For example, if a cloth shop has information about several number of its customers about their yearly frequency of purchases, and the average amount spent per purchase.Then we plot that data to see some kind of a pattern. If the machine can learn this pattern. Then it can make predictions based on what it has learned. Maximum one, two or three dimensions are easy for human brain to understand but machines can learn in many more dimensions like even the thousands.
There are two main machine learning techniques. First one is classification and other one is prediction.
There are three main approaches of learning algorithms. First one is Supervised Learning and second one is Unsupervised Learning. The last one is Reinforcement Learning.
If your train an algorithm with data that also contain the answer within it, then it is called Supervised Learning. For example, if someone train a system to recognize her/his pet by name, he/she must need to identify them for the system.
If you train an algorithm with data where you want the system to figure out the pattern, then it’s Unsupervised Learning for example you might want to input the data about a road map and identify different transportation patterns in that data by itself.
If Someone give any algorithm a goal and expect the machine to achieve that goal in a through try and error scenario, then it is called as Reinforcement Learning. For example, A robot’s attempt to climb over a wall which is difficult to climb until it succeeds.
