Neuronale Netze selbst programmieren: Ein verständlicher Einstieg mit Python - Softcover

Rashid, Tariq

Softcover

ISBN 10: 3960092458 ISBN 13: 9783960092452

Publisher: dpunkt.Verlag, 2024

Synopsis

Neural networks and basics of artificial intelligence are clearly presented

• The bestseller has made a furore because the author explains this both dry and difficult matter exceptionally clearly.
• New in the fully updated 2. Edition: The neural network is finally created with PyTorch to transform it into a typical professional scenario.
• Neural networks are the basis of many everyday applications such as voice recognition, face recognition on photos, the self-driving car, conversion of speech to text etc.

Neural networks are key elements of deep learning and artificial intelligence capable of astonishing today. They are the basis of many everyday applications such as voice recognition, face recognition in photos or the conversion of speech to text. Yet few understand how neural networks actually work.

This bestseller, now in extended 2. Edition, takes you on a fun journey that starts with simple ideas and shows you step by step how neural networks work:

• First, you will learn the mathematical concepts underlying the neural networks. You do not need deeper knowledge of mathematics, because all mathematical ideas are carefully explained with many illustrations. A brief introduction to analysis supports you.
• Then it goes into practice: after an introduction to the popular and easy-to-learn Python programming language, you gradually build your own neural network with Python. They teach him to recognize handwritten numbers until it achieves a performance like a professionally developed net.
• The next step is to improve the performance of your neural network to the point that it achieves 98% number recognition – with simple ideas and simple code. You test the net with your own handwriting and take a look at the mysterious interior of a neural network.
• New in the 2nd Edition: Finally, you create the neural network with PyTorch and transfer it into a typical professional scenario.

Edition details

• Publisher: dpunkt.Verlag
• Publication date: 2024
• Language: German
• ISBN 10: 3960092458
• ISBN 13: 9783960092452
• Binding: Perfect Paperback
• Number of pages: 249