Coding a Full Neural Network from Scratch in Python | Deep Learning

Veröffentlicht am: 01 Januar 1970
auf dem Kanal: Pixagan AI

461

In this lecture, we build a Multi Layer Neural network in python with just numpy as our core dependency.

We start with a visual theory session where we visualize how a neural network works and introduce the idea of a layer in a neural network. Then we visually explore a couple of network architectures

Next, we start with a single neuron we have previously covered but with a tanh activation function.

Finally we manually connect neurons into Layers. Next we build the layers and finally assemble layers into a neural network. We run both the network through a forward and backward pass.

This series is taught by Dr Anil Variyar, who has a PhD in Aeronautics and Astronautics from Stanford University.

Chapters:
00:00:00 Intro
00:01:26 A Layer in a Neural Network
00:06:30 Input, Output and Hidden Layers
00:08:35 Review of Tanh Neuron
00:10:45 Neural Network Example 1
00:17:00 Neural Network Example 2
00:33:45 Building the Neuron in Python
00:42:45 Assembling the Neurons Manually
01:02:30 Creating the Layer Class
01:25:00 Creating the Network Class

Relevant References:

Neural Network Related
Building a Single Neuron Model -    • Coding a Single Neuron Learning Model in P...
Understanding Backpropagation -    • How Backpropagation works | A visual step ...

Relevant Previous Videos
Setting up Anaconda and Jupyter Notebooks : Setting up Python with Anaconda and Using Jupyter Notebooks | A Quick Start

Math Prerequisites
Basics of Derivatives :    • Basics of Derivatives | A Review | Math fo...
What is Chain Rule :    • What is Chain Rule | A Review | Math for AI

#artificialintelligence #deeplearning #neuralnetworks #python #nnfs #machinelearning

Auf dieser Seite können Sie das Online-Video Coding a Full Neural Network from Scratch in Python | Deep Learning mit der Dauer stunde minuten sekunde in guter Qualität ansehen, das der Benutzer Pixagan AI 01 Januar 1970 hochgeladen hat, den Link mit Freunden und Bekannten teilen, dieses Video wurde auf Youtube bereits 461 Mal angesehen und es wurde von 18 den Zuschauern gefallen. Viel Spaß beim Betrachtenden Zuschauern gefallen!

9,539

401