Neural Networks and Deep Learning for Dynamical Systems
Lecture 1
[ View ]
Neural Networks Basics: This lecture provides an introduction to neural networks. A simple classification of dogs vs cats is demonstrated.
MATLAB CODE
Lecture 2
[ View ]
MULTI-LAYER NETWORKS AND ACTIVATIONS: The role of multiple layers and nonlinear activation functions are explored.
Lecture 3
[ View ]
BACKPROPAGATION ALGORITHM: The underlying architecture is formulated for optimizing over the NN weights: Backprop.
Lecture 4
[ View ]
STOCHASTIC GRADIENT DESCENT: The innovation of stochastic into gradient descent is considered.
Lecture 5
[ View ]
DEEP CONVOLUTIONAL NEURAL NETWORKS: The role of convolution and deep architectures is explored for vision applications.
Lecture 6
[ View ]
NEURAL NETWORKS FOR DYNAMICS: This lecture provides an introduction to neural networks and their use for time-series data. A predictor for the Lorenz ODE system is developed.
MATLAB CODE
Lecture 7
[ View ]
NEURAL NETWORK ZOO: This lecture highlights the diversity of neural networks is use in modern data science.
MATLAB CODE