Contact Person : Walton-cara
Phone Number : 15986872308
July 1, 2022
Artificial intelligence lies at the heart of dramatic advances in automotive, healthcare, industrial systems, and an expanding number of application areas. As interest continues to rise, the nature of AI has elicited some confusion and even fear about the growing role of AI in everyday life. The type of AI that enables an increasing number of smart products builds on straightforward but nontrivial engineering methods to deliver capabilities far removed from the civilization-ending AI of science fiction.
Definitions of AI range from its most advanced—and still conceptual— form, where machines are human-like in behavior, to a more familiar form where machines are trained to perform specific tasks. In its most advanced form, true artificial intelligences would operate without the explicit direction and control of humans to arrive independently at some conclusion or take some action just as a human might. At the more familiar engineering-oriented end of the AI spectrum, machine-learning (ML) methods typically provide the computational foundation for current AI applications. These methods generate responses to input data with impressive speed and accuracy without using code explicitly written to provide those responses. While software developers write code to process data in conventional systems, ML developers use data to teach ML algorithms such as artificial neural network models to generate desired responses to data.
How is a basic neural network model built?
Among the most familiar types of machine learning, neural network models pass data from their input layer through hidden layers to an output layer (Figure 1). As described, the hidden layers are trained to perform a series of transformationsthat extract the features needed to distinguish between different classes of input data. These transformations culminate in
values loaded into the output layer, where each output unit provides a value representing the probability that the input data belongs in a particular class. With this approach, developers can classify data such as images or sensor measurements using an appropriate neural network architecture.
Neural network architectures take many forms, ranging from the simple type of feedforward neural network shown in Figure 1 to deep neural networks (DNNs) built with several hidden layers and individual layers containing hundreds of thousands of neurons. Nevertheless, different architectures typically build on an artificial neuron unit with multiple inputs and a single output (Figure 2). Figure 1: Neural networks comprise layers of artificial neurons trained to distinguish between different input data classes. (Source: adapted from Wikipedia)
Figure 2: An artificial neuron produces an output based on an actiation function that operates
on the sum of the nouron's weighted imputs. (Source: Wikipedia)
In a feedforward neural network, a particular neuron n, in hidden layer sums its inputs, x, adjusted by an input-specific weight wp and adds a layer-specific bias factor b (not shown in the figure) as fllows:
Finally, the summed valueS is converted to a single value output by an activation function. Depending on requirements, these functions can take many forms, such as a simple step function, arc tangent, or non-linear mapping such as a rectified linear unit (ReLU), which outputs 0 for S<=0 or s, for S>0.
Although they are all designed to extract the distinguishing features of data, different architectures might use significantly different transformations. For example, convolutional neural networks (CNNs) used in image-recognition applications use kernel convolutions. In this, functions, called kernels, perform convolutions on the input image to transform it into feature maps. Subsequent layers perform more convolutions or other functions, further extracting and transforming features until the CNN model generates a similar classification probability output as in simpler neural networks. However, for developers, the underlying math for popular neural network architectures is largely transparent because of the availability of ML development tools (discussed elsewhere in this issue).Using those tools, developers can fairly easily implement a neural network model and begin training it using a set of data called the training set. This training data set includes a representative set of data observations and the correct casification for each observation- and represents one of the more challenging aspects of neural network model development.
How is a neural network model trained and deployed?
In the past, developers creating training sets had ltte option but to work through the many thousands of observations required in a typical set, manually labeling each observation with its correct name. For example, to create a training set for a road sign recognition application, they
need to view images of road signs and label each image with the correct sign name. Public domain sets of prelabeled data let many machine-learning researchers avoid this task and focus on algorithm development. For production ML applications, however, the labeling task can present a significant challenge. Advanced ML developers often use pre-trained models in a
process called transfer learning to help ease this problem.