China
Africa
Explore chapters and articles related to this topic
Implementation
Published in Seyedeh Leili Mirtaheri, Reza Shahbazian, Machine Learning Theory to Applications, 2022
Seyedeh Leili Mirtaheri, Reza Shahbazian
Microsoft Cognitive Toolkit is a commercial distributed Deep Learning framework with large-scale datasets provided by Microsoft Research. CNTK implements productive DNNs trained for image, speech, handwriting, and text data. Its network is like a symbolic graph of vector operations, namely convolution with building blocks or matrix add/multiply. CNTK supports RNN, FFNN, and CNN architectures and implements stochastic gradient descent (SGD) learning with differentiation and parallelization across multiple GPUs and servers. CNTK is compatible with 64-bit Linux and Windows operating systems using C++, C#, BrainScript API, and Python.
The Coming Cognitive Augmentation Era
Published in Ron Fulbright, Democratization of Expertise, 2020
Naturally, cognitive systems will be the result of artificial intelligence and deep learning software development. The open source ethos is spreading to this domain also. The OpenAI organization was created in 2015 (www.openai.com). OpenAI’s purpose is to promote free and open exchange of artificial intelligence software. The OpenCog project started in 2008 and is an open-sourced artificial intelligence framework. Originally developed by Google, TensorFlow was released in 2015 as an open-sourced machine learning framework. In 2016, the Microsoft Cognitive Toolkit was released as a deep learning framework.
Machine Learning for Solving a Plethora of Internet of Things Problems
Published in Kamal Kumar Sharma, Akhil Gupta, Bandana Sharma, Suman Lata Tripathi, Intelligent Communication and Automation Systems, 2021
Sparsh Sharma, Abrar Ahmed, Mohd Naseem, Surbhi Sharma
Here we introduce the most popular and useful ML tools used in industry and in research. The languages used for ML are Python, R, Javascript, Java, Scala. The most popular framework for implementation of ML algorithms is TensorFlow. It is the most vibrant and useful framework for the implementation of deep learning (DL) models, where neural networks (NN) and computer vision (CV) models are implemented. SciKit-learn is a Python library mainly used for data mining and analysis. A wide range of ML algorithms is implemented on it. Various machine learning algorithms based on supervised, unsupervised and semi-supervised learning types are implemented using Scikit-Learn. Keras is a Python library used for implementing deep neural networks, and it runs on Tensorflow or Theano. PyTorch is a powerful deep learning framework for machine learning written in Python. PyTorch has strong tensor computations similar to the NumPy array with strong GPU acceleration. PyTorch is a tool for ML which is already used on Facebook, Google, Twitter, etc. [81]. Orange is a GUI-based data mining environment used for machine learning and data visualization. It comes with a large toolbox and offers interactive data analysis workflow. It's a drag-and-drop tool for building ML models and analyzing data, and it resides inside the Anaconda Navigator. Google's AutoML is a drag-and-drop method in which various machine learning models can be used without any coding knowledge requirement. Pre-trained neural network models are provided by Google, which is available via its APIs for accomplishing and performing certain required tasks. Graphical user interface of AutoML enables the dragging of image sets by the users. Google Colab is a cloud-based service for ML from Google which supports Python. While using Colab, there is no need for a GPU to be installed in our own system. Colab provides all the GPU-based ML frameworks, including TensorFlow, PyTorch, NumPy, Keras, etc. ANNdotNET is another deep learning tool from Microsoft built on .NET platform. It is a Windows desktop application for creating and training ANN models and is written in C#. The application is considered a GUI tool for the CNTK library, with extensions in data. It relies on CNTK and Microsoft Cognitive Toolkit. Table 18.5 provides the overview of some of the software tools widely used in the implementation of machine learning.
Deep Learning Techniques for OFDM Systems
Published in IETE Journal of Research, 2021
M. Meenalakshmi, Saurabh Chaturvedi, Vivek K. Dwivedi
The Microsoft Cognitive Toolkit called CNTK, an open-source toolkit, was developed for distributed DL. A directed graph describes the computational steps of NNs. In the graph, the operations are represented by nodes, and each edge represents the data flow [5]. CNTK allows to generate and combine DL models like convolutional NNs, feed-forward DNNs, and recurrent NNs [44]. The NN implementation is performed by a specific high-level scripting language, BrainScript.