week3_1

Chapter 6 Slides

Opening Example

 Opening Vignette

 Danske Bank

 Results

 Realize a 60 percent reduction in false positives with an expectation to reach as high as 80 percent.

 Increase true positives by 50 percent.

 Focus resources on actual cases of fraud.

Introduction to Deep Learning

 Deep learning with AI-based learning

Process of developing neural network-based systems  Review Figure 6.11

 Learning process in ANN

 Supervised learning

 Performance function

 Over-fitting

Illuminating the black box of Ann

Deep Neural Networks

Convolution Neural Networks

 Pooling

 Convolution Network unit

Recurrent networks and long short-term memory networks  RNN- specifically designed to process sequential inputs. An RNN basically models

a dynamic system where (at least in one of its hidden neurons) the state of the system (i.e., output of a hidden neuron) at each time point t depends on both the inputs to the system at that time and its state at the previous time point t - 1.

Computer frameworks for implementation of deep learning  Torch

 Caffe

 TensorFlow

 Theano

 Figure 6.36

Cognitive computing

Wrap Up

 Review the Chapter highlights

 Review the key terms

 Complete the weekly homework