Fast multilayer perceptron neural network library for iOS and Mac OS X. MLPNeuralNet predicts new examples by trained neural network. It is built on top of the Apple's Accelerate Framework, using vectorized operations and hardware acceleration if available.
The answer is yes and no. Trying to do deep learning today without access to nvidia gpu is extremely hard. Macs don’t have it today. Even if you buy a pc with an nvidia gpu like a 1080ti, there is a limit to what you can pull with.
- I do use my MacBook for training shallow networks on small training sets. It's good for small prototypes. I use the tensorflow release for CPU. Note that this is specifically true of the MacBook product line. The iMac Pro or Mac Pro lines have very powerful GPU cards, but again, you'll encounter the AMD support challenge.
- Select IAS to connect to the wireless network. If you do not see the AirPort icon in the menu bar. Go to System preferences icon from the dock. Click on Network icon. Select AirPort from the left hand side of the window. Check to see if the status for the AirPort is on. If it is off then click on Turn AirPort On. Select Join Other Network.
Why to choose it?
Imagine that you created a prediction model in Matlab (Python or R) and want to use it in iOS app. If that's the case, MLPNeuralNet is exactly what you need. It is designed to load and run models in forward propagation mode only.
Features:
- classification, multiclass classification and regression output;
- vectorized implementaion;
- works with double precision;
- multiple hidden layers or none (in that case it's same as logistic/linear regression)
Quick Example
Let's deploy a model for the AND function (conjunction) that works as follows (of course in real world you don't have to use neural net for this :)
| X1 | X2 | Y |
|---|---|---|
| 0 | 0 | 0 |
| 1 | 0 | 0 |
| 0 | 1 | 0 |
| 1 | 1 | 1 |
Our model has the following weights and network configuration:
Getting started
This instruction describes on how to install MLPNeuralNet using the CocoaPods. It is written for Xcode 5, using the iOS 7 SDK. If you are familiar with 3rd-party library management, just clone MLPNeuralNet repo on Github and import it to XCode directly as a subproject.
Step 1. Install CocoaPods
CocoaPods is a dependency manager for Objective-C. Installing it is as easy as running the following commands in the terminal:
Neural Networks Pdf
Step 2. Create Podfile
List MLPNeuralNet as a dependenciy in a text file named Podfile in your Xcode project directory:
Step 3. Install MLPNeuralNet
Now you can install the dependencies in your project:
Make sure to always open the Xcode workspace (.xcworkspace) instead of the project file when building your project:
Step 4. Import MLPNeuralNet.h
#import 'MLPNeuralNet.h' to start working on your model. That's it!
Performance benchmark
In this test the neural net is grown layer by layer from 1 -> 1 configuration to 200 -> 200 -> 200 -> 1. At each step the output is calculated and benchmarked using random input vector and random weights. Total number of weights grows from 2 to 80601 accordingly. I understand the test is quite synthetic, but I hope it illustrates the performance. I will be happy if you can propse better one :)
Unit Tests
MLPNeuralNet includes a suite of unit tests in the MLPNeuralNetTests subdirectory. You can execute them via the 'MLPNeuralNet' scheme within Xcode.
Credits
- MLPNeuralNet was inspired by Andrew Ng's course on Machine Learning.
- Neural Network image was taken from Wikipedia Commons
Contact
Neural Network Python
Maintainer: Mykola Pavlov (me@nikolaypavlov.com).
Please let me know on how you use MLPNeuralNet for some real world problems.
License
MLPNeuralNet is available under the BSD license. See the LICENSE file for more info.
Written with StackEdit.