Depth learning framework as Tensorflow and Pytorch provide the API calls available to users, but also hides the underlying implementation details of the depth of learning.
For the convenience of you gain a better understanding of the depth of learning to understand the underlying principles and implementation, is hereby launched the "deep learning theory courses Detailed and Python code to achieve." We expect to "set off your veil, let me see what you look like," depth study further optimization and innovation lay the foundation.
Course Link: https://edu.51cto.com/course/21426.html
This course explains in detail the depth of learning principles and Python code. The program covers the perceptron, the multilayer perceptron, convolutional neural network, recurrent neural network, and using the Python 3 and Numpy, Matplotlib from zero to achieve the above neural network. The course also about the training methods and practical skills neural network, and carry out practical demonstration code. For explanation intensive core, as calculated based on back-propagation algorithm in FIG understood, and mathematical formulas to derive back propagation algorithm; also describes a method for accelerating the convolution im2col.
This course seeks to enable students to learn through in-depth principle, the control algorithm formulas and learn Python code, to get rid of deep learning framework grasp the underlying implementation of the principles and methods.
This course will give participants share the depth of learning Python implementation code. Course code by Jupyter Notebook presentation, runs on Windows, ubuntu and other systems, and without GPU support.
Premise of this course is to use the Python language and Numpy and Matplotlib library.
