Artificial Intelligence

What is Artificial Intelligence (Artificial Intelligence):-

Artificial intelligence is a simulation of the human intelligence process by machines.
In simple words, the ultimate objective of Artificial Intelligence is to build a computer that is exactly like a human brain.
We will explain the meaning of Artificial Intelligence with the help of examples and different types of AI.

Reactive Machines:-

Are extremely common and have been for a long time. These machines do not have their own memory.
They cannot use past experiences to make future decisions. They are programmed to do specific tasks and will continue to do the same thing.
Example- Washing Machine


Limited Memory:-

All the advancements in artificial intelligence are in this category.
They can use past experiences to make future decisions.
For example, spam detection in email is this kind of artificial intelligence.
 The computer understands the types of emails we read,
what is important based on our behavior, and what is spam.
They learn from our past behaviors and react to it in the future.
It is like a human to understand your behavior and react accordingly.
 This is why it is called Artificial Intelligence.
here are two more types of artificial intelligence - Theory of Mind and Self Awareness.
Although they are not currently in practice. The different terms Jaise hears machine learning
 and data science is technologies that are used to manufacture artificial intelligence products.


Deep Learning

Introduction to deep learning
Deep learning is one of the machine learning techniques by which we teach computers what humans are doing.
 For example, driving a car - deep learning plays an important role in driverless car technology in identifying important traffic signs,
 road signs, pedestrian signs, etc. Other major areas of deep learning are voice control in-home systems, mobile, wireless speakers. ,
Alexa, Smart TV, etc. Deep learning for beginners is mostly about the multiple levels of abstraction and representation by which a computer model learns to classify images, sounds, and text, etc. Deep learning models achieve better accuracy and performance than humans in some models... In general, these computer models are trained by a large set of data, labeled and unpublished to identify objects and neural networks that each network has multiple layers.

What is deep learning? 

I am going to explain what is in the word below. Deep learning:
 In general, we do two things all the time consciously or subconsciously,
that is, through our senses (like feeling warm, cold mug, etc.) and prediction. Feel, for example.
 , Predicts future temperatures based on past temperature data. We perform classification and prediction functions for many events or actions in our daily lives such as below: Holding a cup of tea/water/ coffee etc.
 which can be hot or cold. Email classification such as spam / no spam. Classification of daylight such as day or night.
 Future long term planning based on our current situation and the things we have - is called prediction. Each creature in
the world will perform these tasks in its life, for example, consider whether animals like crow will classify a place to build their nest,
 a bee will decide on some factors when and where honey will be found, Bats will come during the night and morning based on day and night
 classification defined Sleep during Let us visualize these functions classification and prediction and they will look similar to
the image below, for classification, we are doing classification between cats and dogs by drawing a line through the data points and in
case of prediction we will be Let's draw a line. Predict when it will grow and decrease.

1)Gradation To classify between cats and dogs in general,
 or between men and women, we do not draw a line in our
 mind and the position of dogs and cats is arbitrary for
the purpose of illustration only and it is useless to say that we
How do you behave in the middle? Our brain dogs are much more complicated than drawing a red line like the one above.
 We will classify between two things based on size, shape, height, form, etc. and sometimes it is difficult to categorize with these characteristics, such as a small dog that has a fury and a new-born cat,
 so it is a clear cut classification. Not in cats and dogs. Once we are children we can classify between cats and dogs,
after that we can classify any dog ​​or cat, even we had not seen it before.


2) prediction For prediction based online, 

we draw through the data points if we can predict where it is most likely to go up or down. The curve is also a prediction of fitting new data points within the range of existing data points i.e. how to close the new data points of the curve. Data points that are in red in the above image (on the right) are examples both within and outside the range of existing data points and try to estimate both curves. In the end, both work classification and prediction end at the same point i.e. draw a shapely line from the data points. If we can train computer models, we can apply it in different models such as drawing a shapely line in three-dimensional planes, etc. to draw a shapely line based on the data points we make. The above thing can be achieved by training a
model with a large number of labels and unlabeled data called deep learning.


Examples of deep learning:

 As we know that deep learning and machine learning are subsets of artificial intelligence,
 but deep learning technology represents the next development of machine learning.
Because machine learning will work based on algorithms and programs developed by humans,
while deep learning learns through a neural network model that performs the same functions as humans and allows a machine or computer to analyze data similar to humans. This becomes possible because we train neural network models with massive amounts of data because data is the fuel or food for neural network models. Below are some examples of deep learning in the real world.

Autonomous vehicle:
Deep learning models are trained with a large amount of data to identify road signs;
Some models specialize in identifying pedestrians, humans, etc. for driverless cars while driving.

Autonomous vehicle:
Deep learning models are trained with a large amount of data to identify road signs;
Some models specialize in identifying pedestrians, humans, etc. for driverless cars while driving.


A career in Deep Education - Introduction

Deep education, called neural organized learning or different levels of education, is a piece of a more comprehensive set of techniques of machine education in
terms of information retrieval of education, rather than special calculations. Learning can be directed,
semi-managed or unsupervised. Deep education offers a further set of systems to enable organizations to take care of complex explanatory issues and drive accelerated development in spurious consciousness. 
By encouraging the calculation of deep learning with a large amount of information, the model can be designed to perform complex undertakings such as undertaking and picture examination. The model of deep learning is identified in almost an organic sensory system with data preparation and correspondence design,
 for example, neural coding that attempts to characterize a connection between different data in the brain and related neuronal responses. Structures of deep learning, for example, Deep nervous system, Deep persistence system and Interdisciplinary neural 
systems PC vision, lecture acceptance, regularundefined Has been associated with areas including bidding management, sound acceptance,
informal community shifting, machine interpretation, bioinformatics and pharmaceutical design, where they have spoken of superior and
 sometimes equally equal to human experts . Careers in deep education are another area of ​​machine education research, presented to draw machine education closer to one of its unique objectives: Artificial Intelligence. This site is expected to classify assets and pointers to data about careers in deep education.


Career Path in Deep Education Deep learning is a standout among today's best-known neural network dialects used as a result of its straightforward image structure, 
and on the basis that it is a universally useful neural programming dialect. You see a career in
 Deep Education used as a part of many fields. New Deep Education engineers have many options for neural programming.
 Be that as it may, careers in deep education alone are not sufficient for the vast majority of these profession options,
they all require supporting abilities. For example, in a system where you need to engage in probabilistic progression with statistics other than neural network systems education. Skills like Convolutional Network, RNN, LSTM, ADAM, Dropout, Batch Norm, Xavier
 / Initialization. A student who is very interested in this profession, they have the practical knowledge on this skill linear regression, SoftMax,
TAN, RELU, Tensorflow. The previously mentioned deep education specialization (AI, undefined Neural Advancement, Data Science and so forth)
each requires a specific height. The software engineer customer gets information assets to execute work obligations in particular application spaces. Data-based analysts in both the scholarly world and industry have given the great case of the Neural Analysis 
Engineer client, however, this gathering is growing in scope. For example, medical specialists (eg, doctors and hereditary instructors)
use the Data Engineer property in pharmacological settings for the motivation behind the analysis, treatment, and patients' advice.
Data Engineer: Researchers are scholars who use computational and artificial techniques with the ultimate goal in mind to inspire a
 logical understanding of living structures. Data Engineer Data Engineer creates novel computational strategies required by customers
and researchers. In this way, a data engineer design should have properties in computational and natural sciences and in biomedical sciences, an undefined Should have the general qualification. Many logical laboratories, both singular custodial academic and vocational divisions, are contracting individuals prepared in deep education to help with laboratory exams. The position is accessible to different levels and types of preparation. Individuals in these positions go into a particular area of ​​research for the most part. Center offices make many organizations a focal asset for laboratories in foundations. These properties are call center offices. Individuals from such gatherings often have a mixture of dependence on various research enterprises with scientists in a wide range of laboratories.


What is the average salary for jobs related to "Deep Education"?
 The average salary for "learn education" is about $ 135,562 per year
for a research scientist for a machine education engineer, $ 135,255 per year.






What is machine learning:


a conversation Many people like us think machine learning is the future.
However, nowadays it is visible in every aspect of our life -
 whether it is Google's computer that plays great Go games or the facility to reply in the inbox of Gmail itself.
 It sounds great to hear, but many people still want to know what machine learning is, or why it matters,
 or why identifying a pet in a photo is not as easy as it sounds. To understand this, we spoke with Maya Gupta,
 a research scientist at Machine Learning at Google. Let's start with the basic questions. What is machine learning?
Machine learning takes some examples to understand their patterns and then uses that pattern to predict new examples in advance.
 Take the example of a film. Suppose that one billion people tell their ten favorite movies. Using these examples, the computer can find out what people like in similar films. Then the computer explains these example undefined With patterns like, "People who like horror movies usually 
don't like romance, but people like films from one actor." Then if you tell the computer that you have Jack Nicholson's film The Shining If liked,
 one can make a guess as to whether you would like Jack Nicholson's 'Something's Gotta Give' as well, and what videos you should suggest on YouTube.
 A little understood. But how does it actually work? Actually, the patterns the machine learns can be very complex and it can be very difficult to explain them in words. Like Google Photos, which lets you search for photos of dogs from all your photos. How does Google do this? First, we gather examples of photos labeled "dog" (thank you to the Internet!). We also submit photos labeled "cat" and millions of others, but I will not talk about it
Then with the help of pixel patterns and patterns of colors, the computer guesses whether it is a cat or a dog (or something). First, the computer guesses which patterns might be good for identifying dogs. It then checks the dog's image used in the example to check if its patterns are working properly. If it is mistaken for a cat as a dog, it makes small changes in the pattern used. Then it looks at the photo of the cat and corrects it again to get the correct pattern.
 And the same process is repeated billions of times - look at an example and if it doesn't look right, improve the pattern used to make that example better.
 Patterns then form models of machine-learning, such as deep neural networks, which can correctly (mostly) identify dogs, cats, firefighters, and many more.

This seems to be the future. What are the other products of Google that currently use machine learning?
Google is doing a lot of new things with machine learning like Google Translate takes pictures of road signs or menus written in a language, finds the words and language in it and translates it to your language right there... 
You can speak almost anything with Google Translate, and speech recognition technology working through machine learning will start its work.
 Speech recognition technology is also used in and products, such as understanding your voice queries in Google app, such as voice queries and making
 YouTube videos more easily searchable.

Our machine learning and artificial intelligence the same?
Actually, these words can mean different things to different people,
but Artificial Intelligence (AI) is broadly a term for computer programs that try to solve problems that humans can easily do,
such as telling about a photo by looking at it. Another work that humans easily do is to learn from examples. And machine learning programs also try to do the same: to tell computers to learn from examples. The interesting thing is that when we understand how to create such a computer
 programs, we can teach them to process a lot of data quickly so that they can also complete difficult tasks like mastering Go (board games), Routing
 all people together in traffic, improving energy usage across the country, as well as my favorite job - finding the best search results for you on Google.

Why is Google suddenly so important to machine learning?
 Machine learning is not a new thing but its roots can be found in 18th-century figures.
 But you are right that a lot of attention is being paid to this recently, for which there are three reasons.
 First of all, we have to gather many examples so that computers can learn to guess better. Then even if it is for those things that you or
 I can do very easily (like finding a dog in a photo). Due to increasing activities on the Internet, we now have a large source of examples from which computers can learn. For example, there are now millions of photos labeled "dogs" on websites around the world, in every language.
 But it is not enough to have too many examples. Now by showing some pictures of dogs to a webcam, you cannot expect to learn anything from them -
 the computer needs a learning program. Recently, more companies and people, including Google, have done encouraging work in this area that shows how complex machine learning programs are and undefined How powerful it can be. Right now our programs are not such that they can be trusted completely,
 as well as much more is yet to be learned for computers, so if there is a slight change in the pattern to get the correct identification, then we have to take many examples - Have to look again. All of this requires a lot of computing power and simultaneous processing. However, there has been so much development in the field of software and hardware that it is possible to do so.

One such thing that computers cannot do today,
but will be able to do soon with the help of machine learning? Until just a few days ago,
the bid recognition technique used to have difficulty in identifying ten credit card numbers on the phone.
 In the last five years, due to machine learning, the technology of speech recognition has developed a lot and now you can use it for Google search. And it is getting better, and faster. I also believe that through machine learning we can look better. Don't know about you, but I don't like wearing them before buying clothes! Due to this, as soon as
 I get a brand of jeans with the right fit, I buy five of its jeans. But with machine learning, we will be able to find out which other brands we can buy by looking at examples of good fitting brands used. Google is not working on it but I hope someone, somewhere, is working on it!

What will machine learning look like in ten years?
One thing that everyone in this field is working on is how to learn fast with fewer examples.
 One way for this (on which Google is especially working very hard) is to give more common sense to our machines, which is called "regularization" in the field. What is meant by giving common sense to the machine? 
One meaning of this is that if an example changes a little, then the machine does not change its decision completely.
 For example, the photo of a dog wearing a cowboy hat should also be placed in a photo labeled dog.
We are teaching machine learning to ignore small, unimportant changes such as cowboy hats to add common sense to learning programs.
This sounds easy to say, but if you made a mistake in it, the machine will not even pay full attention to the important changes! Therefore,
we are trying to bring balance to this work.

What is it about machine learning that makes you most excited?
What inspires you to work on it? I grew up in Seattle,
where I learned a lot about early explorers of the American West, such as Lewis and Clark.
The same spirit works behind researching machine learning that was behind the discovery of those people -
seeing things for the first time, and trying to create a brilliant future from them.
















Comments

Popular posts from this blog

What Is Mechanical Engineering

mechanical engineering diploma