Neural Network is nothing but a Linear Regression

A Towards AI article explains that a neural network without activation functions is mathematically equivalent to linear regression, tracing the modern neural network concept back to multiple linear regression and the simple perceptron. The author details how the equation F(x) = w.x + b mirrors linear regression and notes that activation functions are what differentiate neural networks from regression models.

Neural Networks are a type of Artificial Intelligence AI that are modeled after the human brain. They are used to process large amounts of data in order to make A Neural Network is a type of machine learning algorithm that uses a mathematical model to solve problems. It is similar to Linear Regression in a way that it also uses a set of inputs and weights to produce an output. The difference is that a Regression models are used to identify correlations within the data. Linear regression fits linear equations, like y=mx+c which we studied in algebra. If there is only one variable, this is known as single-variate linear regression. When we use only one variable to in Here in Single Neural Network , y=mx+c ; But it is written as - When there are multiple variables, it is referred to as Multiple Linear Regression or it can be called as Simple Perceptron . The modern Neural Network concept was originated from Multiple Linear Regression . Multiple Linear Regression or Simple Perceptron consists of a single layer of neurons, or nodes, which are connected to other neurons through weighted edges. Each neuron receives multiple inputs from other neurons and produces a single output. The output of the neuron is determined by a linear combination of these inputs, weighted by a set of coefficients, which are learned through a training process. As such, simple perceptron is a type of multiple linear regression, which is a statistical technique for finding the coefficients that best fit a given set of data. Input layer and the layer in the middle hidden layers are densely connected because every input is connected to every hidden unit. Here, we have the input data features , But it is written as - F x = w1 . x1 + w2 . x2 + w3 . x3 + 1 . b So, we can say that Neural Networks are derived from these linear regression techniques and are based on the idea of multiple linear regression models with multiple layers of neurons, or nodes, that are connected to each other in a network. It’s because of the type of activation function. If we use no activation function or linear activation function in Neural Network then the Neural Network becomes a Linear Regression model. STAY WITH ME. I WILL DISCUSS ACTIVATION FUNCTION IN MY ARTICLE. To answer that, we have get our Vector & Matrix concepts clear. Our equation is F x = w.x + b . Here, w = w1 w2 w3 x = x1 x2 x3 To put it simply w = w1 w2 w3 , weight matrix has Now for the Transpose of Active learning is a machine learning technique that can be used to label huge datasets with the help of some labeled data. It involves the use of a model that can be used to predict the label of the unlabeled data and then use the results to select which data should be labeled. By doing this, the model can be trained on a more diverse dataset, improving its accuracy and overall performance. Additionally, active learning can be used to reduce the amount of labeled data needed for training, thus making the process faster and more efficient. Image synthesis is the process of generating a new image or sequence of images using existing ones. It is an important component of computer vision, which is the process of understanding and extracting meaningful information from digital images. Image synthesis can be used to populate a database with new images, create realistic simulations, and generate new visual effects. See you in the next Article. That’s it for now. If you found this article interesting, helpful and if you learn something from this article, please follow, comment and leave feedback. Neural Network is nothing but a Linear Regression https://pub.towardsai.net/lesson-3-neural-network-is-nothing-but-a-linear-regression-e05a328a0f23 was originally published in Towards AI https://pub.towardsai.net on Medium, where people are continuing the conversation by highlighting and responding to this story.