Machine learning | AI
It is the study of algorithms that automatically learns from
the older thing and creates a new path for the designations. It is a subset of artificial
intelligence. It creates sample data to know training data takes place to make
predictions in the future for the response. For simple understanding, a chess
game is the best example. All of them experienced chess played on a computer or
smartphones.
You probably
know that for the last couple of years, Machine learning has become a core
important technology for many products. Also, we know that machine learning is
used as applications in many places. Google, Amazon, Facebook, Microsoft is
using machine learning and investing more for its products like face lock,
search by photo, maps, etc.
Basic fundamental understanding of AI (or Machine learning)
INPUT (DATA) -> PROCESSING (MAKING USE OF THE DATA TO GET
AN OUTPUT) -> OUTPUT (A RESULT BUT NOT JUST PROCESSED ALSO PREDICTED IN
ADVANCE)
When some say about first coming to your mind in the movie Terminator,
Skyline, and I Robots, etc. in which AI is created by humans and the robots are
trying to take over the world. AI also having the positive side and the
negative side. The negative side is there will be huge unemployment and the
positive side is it will be great when in the medical field even we stop Corona
when we predicted the mutation of the COVID-19 virus, all the transport will be
automated and fewer accidents and problems because there will no human errors.
Introduction
By now you have some knowledge about AI. Now about Machine
learning or also called Deep learning. Now let’s see how AI and Machine
learning are related. Since 1956 with the fundamental idea that some complex
intellectual tasks that are performed by humans daily can be also performed by
machines. Those machines can mimic or simulate human cognitive functions such
as learning and complex problem-solving.
Even all
of them are agreed with AI because it cannot performs all the tasks and some machines
in the production line and basic calculator are pre-programmed actually it is not AI.
What is ML?
- When some data is given as an input and output will be of
knowledge to the Machine learning. Let take the chess game as an example, and
look at the above diagram
- When some Data is given as input is processed and given the output
as Answers.
- The Answers are taken as input again with the Data and new
Rules are created.
- This is learned from the earlier mistakes and moves in the
chess game.
- When this goes on Machine leaning improves everything and every
time.
- The training box is called Black Box and the process also
known as Teaching.
- In simple words, it is learning from the data without being
programmed with a pre-set of rules and Adapting the Knowledge.
Deep learning
Imagine that the ML knowledge or also called the trained
model that was created during the training phase, has some size or capacity, like
a small brain, medium brain, very large brain. In Machine learning, we are not
using the small, medium, or large brains to describe the capacity of a trained
model. It is described using the concept of layers. A layer is the basic
building block of deep learning.
A single layer is like a data transformation phase and you
can have several layers that are connected, so the output from one layer can be
the input to another layer. So each layer receives an input transform it with an
asset of mathematical functions and then it will pass those values to the next
layer and then it will propagate until reaching the final layer.
If we put very small amounts of layers inside the ML system
like one or two, it is called Shallow
learning. The learning algorithm will catch a relatively small amount of
patterns while learning from the data which can be more than enough for a
specific use case. On the other hand, if we put more hidden layers in our
machine learning box, in our model, there is a chance to collect more
knowledge, more patterns from the data. This is called Deep learning, and it is referring to the simple fact that more
layers are used inside. Those layers in deep learning are also called neural
networks as they in some level inspired by biological neural networks as we
have humans and animals. Well, today there are systems with thousands of
layers. As you may guess a bigger brain can handle much more complex tasks,
deep learning is probably the most exciting field under machine learning.
Applied AI vs. Generalized AI
The existing industry implementation of machine learning is
focused on performing very narrow tasks. This is the input and this is the output,
meaning each ML use case is used to accomplish a single specific task. Those
kinds of AI use cases are called “Weak AI” or also called “Applied AI” as they
are focused on narrow tasks that can be applied in practical applications.
When it comes to extremely complex and almost impossible
today to create an AI machine that will perform multiple tasks like the human
brain. This is called “Generalized AI ” or also called “Strong AI” meaning the
intelligence of a machine that can understand and learn almost any intellectual
task that a human being can learn to perform.
Why now?
The ingredients making AI flourish everywhere. Data is the
first ingredient. Data is the raw material consumed by the ML machine. Today
more than ever data is available from a growing amount of data sources. Every action
performed on electronic devices like your mobile phone is recorded and stored.
Every action we perform on some social network is recorded, when we tag a
picture on Facebook we just creating some data on that picture. This is called
user-generated data.
The second ingredient for machine learning is hardware. The hardware
is important when handling large data sets. Inside any computer, there is a
model called the CPU, that is most of the calculations which need to be a
powerful one. On the hand GPU, this is a dedicated high-performance chip to handle fast graphics,
data manipulations
