Downloaded 44 times
Artificial Intelligence = ML + DL with Tensor Flow
The document provides an overview of artificial intelligence (AI), its history, applications, and methods including machine learning and deep learning, using TensorFlow as a key example. It details the differences between supervised and unsupervised learning, classification versus prediction, and the functionalities of TensorFlow developed by Google. Additionally, it lists companies utilizing TensorFlow and mentions its capabilities in handling complex computations across multiple platforms.
More Related Content
Similar to Artificial Intelligence = ML + DL with Tensor Flow
More from Amity University | FMS - DU | IMT | Stratford University | KKMI International Institute | AIMA | DTU
Artificial Intelligence = ML + DL with Tensor Flow
- 1. Artificial Intelligence = ML +DL with Tensor Flow Dr. Kamal Gulati
- 2. Contents • Artificial Intelligence •History of Artificial Intelligence • Artificial Intelligence Use Cases • Artificial Intelligence Applications • Ways of Achieving AI • Machine Learning • Deep Learning • Supervised and Unsupervised Learning • Classification Vs Prediction • TensorFlow • TensorFlow Graphs • History of TensorFlow • Companies using TensorFlow • Using Deep Q Networks to Learn Video Game Strategies • TensorFlow Use Cases • AI & Deep Learning with TensorFlow • How TensorFlow used today
- 3. Sci-Fi Movies thatmight Become Reality
- 4.
- 6. What is ArtificialIntelligence?
- 7. History of ArtificialIntelligence
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16. Ways of Achieving AI MachineLearning1 Deep Learning2
- 17. Machine Learning Machine learning refersto the use of algorithms to parse data, process and learn from it, in order to make predictions or determinations about something. One of the best application for machine learning is computer vision: OCR, object tracking, object recognition etc.
- 20. Deep Learning Deep learning isa subfield of machine learning concerned with algorithms inspired by the structure and function of the brain called artificial neural networks (ANNs) Compared to older ML algorithms, Deep Learning performs better with a large amount of data
- 22.
- 23. Machine Learning andDeep Learning
- 24. Supervised and UnsupervisedLearning Supervised Learning • All data has been labeled (supervised) by an expert. Thanks to this labeling process, we can help the network to realise the difference between classes (even though sometimes this does not happen). • Some techniques: NNs, SVM, etc. Unsupervised Learning • Our data are not labeled. Unsupervised algorithms consider confidence measures among samples in order to create homogeneous clusters. • Most famous technique: Clustering (k-means, hierarchical etc.)
- 25. Classification Vs Prediction Classification •Given an input observation, classification is the problem of identifying to which of a set of categories (classes) the new observation belongs. • i.e: trafficsignals recognition, emotion recognition etc. Prediction • Prediction refers to the problem of estimating the behaviour of a phenomenon by analysing the “previous history” • i.e: object tracking, forecasting etc.
- 26.
- 27. TensorFlow TensorFlow is anopen-source library for numerical computation and machine learning. Its basic principle is simple: you build in Python a graph of computation to perorm and then TensorFlow runs it efficiently using optimized C++ code. TensorFlow supports computation across multiple CPUs and GPUs
- 28. TensorFlor Originally developed byGoogle Brain Team to conduct machine learning and deep neural networks research. General enough to be applicable in a wide variety of other domains as well . Provides an extensive suite of functions and classes that allow users to build various models from scratch. https://www.youtube.com/watch?v=MotG3XI2qSs
- 29. Tensor Flow Graphs Runninga Simple Graph • Software that uses TensorFlow is often divided into two phases: graph building and execution. • In order to evaluate this graph we must run the session and all its initialisers. • TensorFlow supports computation across multiple x = tf.Variable(3,name=”x”) y = tf.Variable(4,name=”y”) f = x*x + 2*y + 5 sess = tf.Session() sess.run(x.initializer) sess.run(y.initializer) res = sess.run(f) print res x = tf.Variable(3,name=”x”) y = tf.Variable(4,name=”y”) f = x*x + 2*y + 5 with tf.Session() as session: x.initializer.run() y.initializer.run() result = f.eval()
- 30. History Google Brain's second-generationsystem. Version 1.0.0 was released on February 11, 2017. can run on multiple CPUs and GPUs. Available on 64-bit Linux, macOS, Windows, and mobile computing platforms including Android and iOS.
- 31.
- 32. Companies using TensorFlow • Google •OpenAI • DeepMind • Snapchat • Uber • Airbus • eBay • Dropbox • A bunch of startups
- 33. Using Deep QNetworks to Learn Video Game Strategies
- 34.
- 35. AI & DeepLearning with TensorFlow
- 36. AI & DeepLearning with TensorFlow
- 37. AI & DeepLearning with TensorFlow
- 38. AI & DeepLearning with TensorFlow
- 39. AI & DeepLearning with TensorFlow
- 40. How TensorFlow usedtoday?
- 41. For more updateson Big Data, Cloud Computing, Data Analytics, Artificial Intelligence, IoT subscribe to http://www.mybigdataanalytics.in