Tensor+Flow=TensorFlow

-

 


With the increase in popularity of machine learning and deep learning among students, the term TensorFlow has also been in demand. So today I am here to tell you briefly about what Tensorflow is, How it is implemented, and why it is important.

Brief history

Tensorflow was developed by Google Brain and it was released under Apache 2.0 license in November 2015. The current stable version of TensorFlow is 2.3.1 and it is a popular GitHub repo with 152k plus stars. TensorFlow is a vibrant, active community of developers with more than 2000 developers actively contributing to the code base. 


What is Tensorflow?


Tensorflow is an end-to-end open-source platform for machine learning.

Tensorflow name itself signifies a meaning, It is made up of two words: tensor and flow. A tensor is the multidimensional array and flow is a graph of operations. 

Internally, TensorFlow uses machine learning algorithms as a working graph in a multidimensional array. Once you have accessed the data in TensorFlow, there is a calculation that needs to be done. Each calculation in TensorFlow is presented as a Data Flow Graph. It is not the same as traditional programs. We prepare graphs with nodes and then they are executed in the form of a session with the data in the Tensors. Each node in the graph represents mathematical performance (add, subtract, multiply, divide, etc.) and each edge represents the multi-item layout that is Tensors. A graph is created and data are processed.


Its Importance?

TensorFlow provides easy to build and deploy machine learning models for a newcomer in machine learning. 
If you are a machine learning researcher, TensorFlow enables you to build state-of-art machine learning models with Keras functional API and model subclassing APIs.
TensorFlow API is available for Python, Java, and Go programming languages.
TensorFlow has a very flexible architecture. It enables easy deployment across different hardware platforms like CPUs, TPUs, and GPUs, and computing devices like desktops, servers, mobile devices, and edge devices.

Tensorflow Everywhere


Google uses TensorFlow to improve its products such as Gmail or doc. Airbnb, for example, uses TensorFlow to separate images and find objects in their set of images. AIRBUS uses TensorFlow to find interesting content on satellite imagery and make it available to its customers. TensorFlow is also used for many good social and financial platforms as PayPal uses TensorFlow to get fraudulent transactions, Twitter uses TensorFlow to use tweets. Therefore, you can see that TensorFlow is a flexible product and is used to develop and export machine learning models by companies in various domains.

You can also check out some of these case studies on the tensorflow.org website.



Isha Mudgal (2018-2022)

Computer Science Department
KIET Group of Institutions

Location: Monthly Wrap KIET Ghaziabad, Uttar Pradesh 201206, India

Comments

Post a Comment

Popular posts from this blog

QUANTUM STORAGE & MEMORY

-
Image
  Quantum computers stand to transform the world of computers. They have the capacity to be greatly more powerful than our current digital devices, with the power to solve important computational problems and simulate complex physical systems with unmatched efficiency. Like memories are an important component for computers, that’s like quantum memories are also essential components for quantum computers, a new generation of data processors that follow the laws of quantum mechanics and these laws help to overcome the limitations of classical computers. Quantum computers are expected to be much faster and more powerful than their traditional analogue as information is calculated in qubits (the fundamental unit of computation in quantum computing)  in quantum computers while the older units are bits which used in classical computers, can represent both 0 and 1 at the same time. Quantum memory is used as a key element in information processing, such as optical  quantum comp...
Read more

QUANTUM COMPUTING: CAN FIGHT CLIMATE CHANGE

-
….IN AN INFINITE UNIVERSE ANYTHING THAT COULD BE IMAGINED MIGHT SOMEWHERE EXIST.                                                                   - Dean Koontz Quantum computing is an emerging domain that is yet to enter our lives. However, the researchers are working hard to bring it into existence. Quantum computing can make the impossible possible. There are some problems that even supercomputers can’t solve. Quantum computers will however have the potential to solve such problems. In the same way that bits are used in a classical computer, quantum computers use quantum bits or qubits which can store information in the quantum form. A qubit can be 0 or 1, or ...
Read more

BLOCKCHAIN: HOW IT WORKS AND WHY SO POPULAR

-
Image
Merely, 12 years ago on 31 st October 2008, SATOSHI NAKAMOTO published a paper on the internet about the idea of bitcoin. The Paper’s first few lines almost depicted all. The lines – “A version of electronic cash that would allow payments to be sent directly from one party/person to another party without involving a financial institution.” The value of bitcoin in 2010 was ~ 0 INR and value in 2020 ~ 15 lakhs INR. Recently, it touched its high price point, and now again is the topic of discussion among the market and media. Why so much difference? Like the US dollar is controlled by the Central Bank of US and the Indian rupee is controlled by RBI. Normally, our financial systems are based on “TRUST”. Whereas Cryptocurrency is a digital asset over which central banks or financial institutions have no control or rules/regulation. The currency notes and coins have value just because they are guaranteed by the government and central bank. Every currency has a promissory note with the signa...
Read more