Practical Deep Learning is a 3 day hands-on instructor led training class that will enable students with no Artificial Intelligence knowledge to understand the basics principles of AI and Deep Learning and apply that knowledge to practical problems.
The course is 50% lecture and 50% hands on code writing, debugging, and testing on reasonable sized examples. The lecture will describe the AI techniques needed create different AI models and the labs will reinforce those techniques with hands on usage.
The examples used in the labs can be extended for real projects.
Who should attend?
- Engineers who wish to understand the hows and whys of AI techniques and models
- Engineers who need to learn Deep Learning techniques for application to an existing or future project
- Engineers who have a rudimentary AI understanding but want to fill in their knowledge
Structure and Content
Introduction to different types of AI models, and the focus of the class.
Linear Regression definition, how to create models, how to train linear regression models, how to generate data, how to test accuracy of regression models
Classification definition, basic neuron definition and operation, neuron creation, neuron training, accuracy testing
Classification using multiple neurons, multiple neuron models, multiple neuron error calculation, multiple neuron optimization, multiple neuron training
Deep Neural Networks
Activation functions, multiple layer network creation, multiple layer operation, multiple layer optimization, multiple layer error propagation, loss differentiation
Introduction to AI frameworks, Introduction to Tensorflow
Tensorflow Linear Regression
Placeholders, Variables, Sessions, Running sessions, optimization, calculating error
Softmax definition, Cross Entropy definition, error propagation, Tensorflow Optimizers, Learning rate, epochs
Tensorflow Deep Networks
Tensorflow activation functions, Tensorflow differentiation, Tensorflow error propagation, Tensorflow Deep Network optimization, Deep Network descriptions
Visualizing Model Operation
Tensorboard, adding summaries, adding histograms, adding graphs, interpreting results
Convolutional Neural Networks
Convolutional filters, feature maps, convolutional layers, pooling layers, fully connected layers, stride, padding, constructing CNN networks, training CNN networks
What makes a good data set, balanced data sets, distinct data sets, non-conflicting data, ImageNet, Inception-V3, transfer learning description, transfer learning operation, transfer learning
TensorFlow Lite, weight quantization, operation on mobile/embedded devices