.
"Artificial Intelligence"@en .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
"Using Python and Tensorflow for scientific engineering and artificial intelligence calculations"@en .
.
.
"https://www.kedivim.tuc.gr/el/ekpaideysi/ola-ta-programmata/chrisi-python-kai-tensorflow-gia-epistimonikoys-ypologismoys-michanikis-kai-technitis-noimosynis"@en .
"NaN" .
"12.0" .
"Presential/Online/Hybrid"@en .
"Acquisition of basic programming skills with the python programming language with emphasis on the use of specialized packages.\nAcquisition of basic knowledge of artificial neural networks and their implementation using specialized packages such as tensorflow.\nApplications of artificial neural networks in engineering and related sciences using the aforementioned tools.\n\nIn detail, the program includes the introduction to the Python programming language and the specialized artificial intelligence package Tensorflow with examples from solving engineering problems and applications of neural networks in solving engineering problems and partial differential equations. He also has examples of finite element problem solving, using neural networks to solve parameter identification problems and solving differential equations through physics-informed neural networks.\n\nFirst, the basics of the Python language will be presented (variables and data types, arithmetic and logical operations, basic built-in functions, program structure, data structures). Also, all the necessary data on algorithmic structures (program execution flow, sequence structure, selection and iteration structures), functions (creation, call and parameterization) will be presented.\n\nThe project will present the basic concepts of artificial neural networks (structure and components, artificial neuron, Perceptron, weights and synapses, network architecture, activation functions, properties, advantages and uses). The various training methods are also examined. Specifically, supervised training (perceptron, back propagation) and unsupervised training (Hebb's algorithm, competitive learning, Kohonen networks) are examined. Solving differential equations through neural networks. According to this approach, the network is trained to solve problems governed by general nonlinear partial differential equations. Depending on the nature of the available training data, the models are divided into continuous and discrete time.\n\nFinally, the Tensorflow package will be imported using Tensorflow's Python API. In order to better understand the concepts involved, an important part of the Tensorflow practices and tools will be covered through relevant examples and exercises.\n\nModules\nProgramming principles and basics of the Python language\nArtificial neural networks and implementation with tensorflow\nApplications of artificial neural networks in engineering. Processing direct problems.\nSolving inverse problems (parameter identification) in engineering" .
"3" .
"NaN"@en .
.
.
"Technical University of Crete"@en .
"NaN"@en .
"no data" .
"no data" .
"158"^^ .
"no data"@en .
.
"Greek"@en .
.