Untitled

Train your own handwritten digit recognizer in C# .NET

Convolutional neural networks (CNNs) are the current state-of-the-art model architecture for image classification tasks. CNNs apply a series of filters to the raw pixel data of an image to extract and learn higher-level features, which the model can then use for classification. We have seen in the earlier post about convolution neural network in the two post as mentioned below:

https://www.te…

GPU Accelerated programming with Amplifier

With the thirst to build sophisticated AI or to build complex algorithm to solve a problem, the need for faster code execution is always a challenge. For many years, GPUs have powered the display of images and motion on computer displays, but they are technically capable of doing more. Graphics processors are brought into play when massive calculations are needed on a single task. GPGPU…

Introducing Arithmetica – A simple but useful math library with Quantum Computing Simulator

Arithmetica is another hobby project started in love with Maths and Quantum Physics. It’s strange I use to run from these maths and physics subjects in school but now I think I find it more interesting when I am able to understand the simple and humbleness behind it. It always looks daunting when we talk about these subject but trust me it’s as easy as learning any language, only thing you have…

Learn Coding Neural Network in C#: The convolution function

This post is in continuation of the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

Last post, I discussed the human visual cortex and mapped to the convolution function of it. Convolution is the most crucial building block in the computer vision and once you have the understanding of it, you will be in love with…

Learn Coding Neural Network in C#: Understanding The Visual Cortex

This post is in continuation of the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

So far we were implementing and seeing the application of the fully connected layers which is considered as a simpler and generalized version of the human brain. Although this architecture is capable of solving complex problems when it…

Learn Coding Neural Network in C#: Predict Boston Housing Prices

This post is in continuation of the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

In this post, we will take a real life regression problem. We will take the Housing dataset which contains information about different houses in Boston. This data was originally a part of UCI Machine Learning Repository and has been…

Learn Coding Neural Network in C#: Solve Titanic Survival problem

This post is in continuation of the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.  We are going to use the library we build from scratch to solve some complex problem, we are taking Titanic Survival classification as our first example.

The sinking of the RMS Titanic is one of the most infamous shipwrecks in…

Learn Coding Neural Network in C#: Build Sequential Model with Training and Prediction

This post is in continuation to the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

In this post, we will nearly finish with our own deep learning library developed from scratch which will be capable of training and predicting. This will be a last extensive stretch of coding where we are going to develop an API type…

Learn Coding Neural Network in C#: Adam Optimizer to correct the network

This post is in continuation to the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

So far in the series we have defined the layers, activation, cost and metrics with forward and backward function. We have also seen how the gradient is calculated for each layer parameter which will be used to update the neural…

Learn Coding Neural Network in C#: The backpropagation technique – Part 2

This post is in continuation of the back-propagation technique implementation in C# project. https://www.tech-quantum.com/learn-coding-neural-network-in-c-the-backpropagation-technique-part-1/ .

If you are not sure where to start, please go through this post first.

Last post we implemented backward method for layers and activation function. Just to recall the back-propagation method is…

Learn Coding Neural Network in C#: The backpropagation technique – Part 1

This post is in continuation to the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

In this post we will implement the back propagation to calculate gradients of the layers parameters so that they can be optimized by an optimizer function. In the examples so far we were using 2 fully connected layers passed on to the…

Learn Coding Neural Network in C#: Define Loss/Cost and Metrics for the neural network

This post is in continuation of the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first. By the end of this post, we will learn to build loss/cost and metric functions which will help fine-tune and measure the model performance.

Cost Functions

Its a method of evaluating how well your neural network model fit with…

Learn Coding Neural Network in C#: Define layers and activations

This post is in continuation to the learning series “Learn Coding Neural Network in C#”. If you are not sure where to start, please go through this post first.

By the end of this post you will be ready for a framework with Layers and Activation function defined. The code will be well structured to be able to extend to more layers and functions.

Some background: A layer is the…

Learn Coding Neural Network in C#: Build your own Tensor with Math Ops

When we represent data for machine learning, this generally needs to be done numerically. Especially when referring specifically to neural network data representation, this is accomplished via a data repository known as the tensor. A tensor is a container which can house data in N dimensions. Mathematically speaking, tensors are more than simply a data container, however. Aside from holding…

Learn Coding Neural Network in CSharp: Understanding what we are going to do

Long time back I wrote blogs about explaining the basic function of a neuron along with a mathematical explanation on how to represent a neuron. Explained how to represent the neuron in the form of vectors and matrix. Then I wrote a very simple neural network to explain the working of a bunch of connected neurons to solve a simple problem. If you follow the below blogs first you will understand…

Learn Coding Neural Network in CSharp: Understanding what we going to do

Logo time back I wrote article about explaining basic function of a neuron along with mathematical explanation on how to represent a neuron. Explained about how to represent the neuron in form of vectors and matrix. Then I wrote a very simple neural network to explain the working of bunch of connected neurons to solve a simple problem. If you follow the below blogs first you will understand what…

Speed up neural network training with batch normalization

Sometimes training neural networks can take days or weeks. This has several disadvantages and has constantly troubled machine learning practitioners, especially those without access to more computing resources. Long training time often means that you cannot tune models until they are done training – several days or weeks. Even after they are done training, you cannot properly explore all possible…