Neural Network Tutorial - Ch. 10.3 Example: Parabola fitting

Показать описание

In this video we train a back propagation network to do some simple curve fitting, not using discrete data points as source, but instead generating them from a function. Conceptually, this is easily extensible to training a network to learn something based on user input, or based on a simulation, which itself may be quite complex, but whose solution may be approximated by the network.

Additionally, a brief discussion on training data quality and domain is presented, displaying how you must be careful not to extrapolate with BPNs.

Рекомендации по теме

Комментарии

Hi, my neural network is not converging with ease on this function. I'm wondering if your neural network is configured a great deal different than mine. Thank you. My configuration:
layerSizes = {1, 2, 1}, with biases on all layers, and the first two layers activated with the Tanh function. The last layer has no activation function applied to it. I am not using batch normalization.

jakeharrison

Yea, I wasn't about to go into using hyperbolic trig functions... You can also add the arc-tangent, but both trig and exponential take more time to compute than rational expressions. (Not to say you couldn't use it)

TheRyanFellow

Initializing the network with 10 nodes in the hidden layer, it got a result at ~96K epochs. :)
Not consistently though. ;)
This one seems to get it most of the time: int[] layerSizes = new int[4] { 1, 20, 20, 1 };

sorentauberlassen

Thanks! Unfortunately I have never done anything with time-series prediction so I wouldn't want to try and teach it. It is interesting though, and conceptually seems simple. As far as the csv file, I assume you want that in the context of loading time-series data. Couldn't you just open the file and parse it? I've never actually tried... haha

TheRyanFellow

Neural Network Tutorial - Ch. 10.3 Example: Parabola fitting

But what is a neural network? | Chapter 1, Deep learning

Neural Network In 5 Minutes | What Is A Neural Network? | How Neural Networks Work | Simplilearn

Explained In A Minute: Neural Networks

The Essential Main Ideas of Neural Networks

Neural Networks Explained in 5 minutes

How to Create a Neural Network (and Train it to Identify Doodles)

Neural Network Simply Explained | Deep Learning Tutorial 4 (Tensorflow2.0, Keras & Python)

Neural Network Tutorial - Ch. 4.1 The constructor (part 1)

Neural Networks & Image Recognition: How AI Understands Pictures in 2 mins

Gradient descent, how neural networks learn | Chapter 2, Deep learning

What is backpropagation really doing? | Chapter 3, Deep learning

Neural Network Full Course | Neural Network Tutorial For Beginners | Neural Networks | Simplilearn

All Convolution Animations Are Wrong (Neural Networks)

Neural Networks explained in 60 seconds!

Tutorial 2- How does Neural Network Work

Create a Basic Neural Network Model - Deep Learning with PyTorch 5

Neural Network Tutorial | Artificial Neural Network Tutorial | Deep Learning Tutorial | Simplilearn

Neural Networks Explained - Machine Learning Tutorial for Beginners

TensorFlow 2.0 Complete Course - Python Neural Networks for Beginners Tutorial

Tutorial 4: How to train Neural Network with BackPropogation

Neural Network Crash Course 2024| Neural Network Tutorial For Beginners |Neural Network|Simplilearn

Simple explanation of convolutional neural network | Deep Learning Tutorial 23 (Tensorflow & Pyt...

Deep Learning | What is Deep Learning? | Deep Learning Tutorial For Beginners | 2023 | Simplilearn

Keras with TensorFlow Course - Python Deep Learning and Neural Networks for Beginners Tutorial