China
Africa
Explore chapters and articles related to this topic
Artificial Intelligence
Published in Ravi Das, Practical AI for Cybersecurity, 2021
The second sub-field next to be examined is that of the Neural Networks (also known as NNs). A specific definition for it is as follows:Neural networks are a set of algorithms, modeled loosely after the human brain, that are designed to recognize patterns. They interpret sensory data through a kind of machine perception, labeling or clustering raw input. The patterns they recognize are numerical, contained in vectors, into which all real-world data, be it images, sound, text or time series, must be translated.Neural networks help us cluster and classify. You can think of them as a clustering and classification layer on top of the data you store and manage. They help to group unlabeled data according to similarities among the example inputs, and they classify data when they have a labeled dataset to train on. (Neural networks can also extract features that are fed to other algorithms for clustering and classification; so you can think of deep neural networks as components of larger machine-learning applications involving algorithms for reinforcement learning, classification and regression).(Pathmind, n.d.)
Computational Intelligence
Published in Yeong Koo Yeo, Chemical Engineering Computation with MATLAB®, 2020
Artificial neural networks are a set of algorithms, modeled loosely after the human brain, that are designed to recognize patterns. They interpret sensory data through a kind of machine perception, labeling or clustering raw input. The patterns they recognize are numerical, contained in vectors into which all real-world data, be it images, sound, text, or time series, must be translated. Neural networks help us cluster and classify. They help group unlabeled data according to similarities among the example inputs, and they classify data when they have a labeled data set to train on. Neural networks belong to a group of information processing techniques which can be used to find knowledge, patterns, or models from a large amount of data.
Current Technologies
Published in Ravi Ramakrishnan, Loveleen Gaur, Internet of Things, 2019
Ravi Ramakrishnan, Loveleen Gaur
Here, AI refers to intelligence demonstrated by machines. However, as things become more mainstream, they get removed from the purview of the AI such as optical character recognition is no longer considered part of AI. Recent advancements have led to the evolution of self-drive cars and drones and deep learning is further helping evolve trends from machine datasets. These datasets may be structured like text and databases or unstructured like images and videos and have the potential to be an important source of knowledge from the mid-twentieth century (McCarthy, 1959; McCarthy, 1977). With statistical methods and libraries, there have been significant developments of applications (Weikum et al., 2009). The objective of AI is to analyze the situation and take actions with a probability of increasing its goal of successful outcomes. These goals can be either predefined or goals can be derived from training data such as in the K-Nearest Neighbor algorithm. Some AI algorithms are self-learning from data and the objectives are self-planning and learning through automation. The next generation of AI based on natural language processing confers on machines the ability to read and write in human language (Russell & Norvig, 2009). The concept uses, other than word matching, lexical meaning matching to group data and categorizing searches. Machine perception or computer vision is yet another category that can use sensor inputs such as camera images, voice, and video and it is used for face recognition or speech recognition. AI has predominant usage in robotics, and assists in movement based on experience and can navigate shop floors with different layouts. AI can work on top of the IoT generated machine data to make machines more autonomous and there are numerous examples integrating both of these technologies. iRobot Roomba, a vacuum cleaner, has AI to map home locations and adapt to different surfaces or new items. Apple Siri is an AI embedded inside an internet-connected mobile phone device. Nest Labs thermostat solutions, which are IoT devices, not only captures room temperature and control but also uses AI to learn about user preferences and accordingly suggest them. Tesla Motor electrical cars operates as a fleet network and all the cars share collective learnings as part of AI. This consists of connected sensors to generate raw data for further analysis.
Towards a framework to design product service system-based mobile phone waste management: A social media data analysis perspective
Published in International Journal of Computer Integrated Manufacturing, 2023
Mohadeseh Pourabbasi, Sajjad Shokouhyar
Where defined as the current time step; is the weight vector indicates the location of an output unit in the data space at a time ; represents the reducing neighbourhood function for the training time and the distance from ; is the input vector generated from the input data at the time . Thus, with the applied learning process, the high-density regions in the input space attract the weight vectors, grouping them into clusters according to their distance. SOM provides valuable information on the system according to the clusters shaped in the case of the weight vectors’ attraction to the high-density areas. Thus, SOM desires to fulfil sophisticated works, including control, process analysis, and machine perception.
Stick-Slip Classification Based on Machine Learning Techniques for Building Damage Assessment
Published in Journal of Earthquake Engineering, 2022
Yunsu Na, Sherif El-Tawil, Ahmed Ibrahim, Ahmed Eltawil
To the author’s knowledge, no prior studies have examined stick-slip motion classification based on machine learning. Stick-slip motion classification is analogous to Human Activity Recognition (HAR) as both use smart phone generated acceleration data to discover meaningful characteristics. For the latter, machine learning algorithms using accelerometer readings have been successfully used to detect the type of activity. The accelerometer data used are typically categorized into separate classes through a classification process. This classification problem is a multidisciplinary research area which shares connections with machine learning, machine perception and ubiquitous computing. Because of its wide-ranging capabilities, classification using accelerometers has drawn increasing interest from researchers in a variety of fields.
Two-dimensional coronene fractal structures: topological entropy measures, energetics, NMR and ESR spectroscopic patterns and existence of isentropic structures
Published in Molecular Physics, 2022
Micheal Arockiaraj, Joseph Jency, Jessie Abraham, S. Ruth Julie Kavitha, Krishnan Balasubramanian
As fractal systems become larger as a function of their dimension, one has to consider a combinatorial library of fractals generated from such building units as well as complex metal organic frameworks. For such a large library, machine learning techniques combined with artificial intelligence are becoming vital in many areas of molecular sciences including drug discovery [70–72]. Consequently, it is important to develop algorithms that are suitable for the machine perception of molecular structures through rapid computations of their spectra such as NMR spectra and ESR hyperfine patterns. Hence we consider the two isentropic pairs of fractals of coronene, namely, and (see Figure 5). As a first step towards the machine generation of NMR or ESR spectral patterns, we outline a few combinatorial and graph theoretical techniques.