China 
Africa 
Explore chapters and articles related to this topic
Basic Approaches of Artificial Intelligence and Machine Learning in Thermal Image Processing
Published in U. Snekhalatha, K. Palani Thanaraj, Kurt Ammer, Artificial Intelligence-Based Infrared Thermal Image Processing and Its Applications, 2023
U. Snekhalatha, K. Palani Thanaraj, Kurt Ammer
Feature extraction is an essential step in the machine learning process, wherein useful data and measurements are derived from the set of images that are informative, non-redundant, and help to train the machine learning models with a much better performance. In deep learning models, feature extraction layers are often associated with the reduction of dimensionality. There are various algorithms that may be used to extract useful features from the picture, some of which have been listed below:Gray Level Co-occurrence Matrix (GLCM)Speeded-Up Robust Features (SURF)Gray Level Run Length Matrix (GLRLM)Local Binary Pattern (LBP)Gray Level Size Zone Matrix (GLSZM)Local Binary Gray Level Co-occurrence Matrix (LBGLCM)Scale-Invariant Feature Transform (SIFT)Local Directional Pattern (LDP)Segmentation based Fractal Texture Analysis (SFTA)
Brain–Computer Interfaces (BCIs)
Published in Teodiano Freire Bastos-Filho, Introduction to Non-Invasive EEG-Based Brain–Computer Interfaces for Assistive Technologies, 2020
Alessandro Botti Benevides, Mario Sarcinelli-Filho, Teodiano Freire Bastos-Filho
BCI is the union of two main processes: signal acquisition and signal processing (Figure 2.1). Regarding the signal acquisition, it is important to define the standard that is used for the placement of the electrodes on the scalp and the sampling rate of the system. The signal processing stage comprises signal preprocessing, feature extraction, pattern classification, and translation of the mental task to commands for a BCI application. The signal preprocessing stage is intended to reduce the amount of noise that contaminates the EEG signal, which usually employs spatial filters or high-order statistical (HOS) separation methods. The feature extraction stage focuses on finding the main features that differentiate the mental tasks, which can be followed by a feature selection step to refine the search for the best features. Then, the EEG features are sent to a classifier whose output is translated into commands to a BCI application.
What Is Data Analytics?
Published in Rakesh M. Verma, David J. Marchette, Cybersecurity Analytics, 2019
Rakesh M. Verma, David J. Marchette
A measurement is a value extracted from some object that one is investigating. For example, in health care this might be blood pressure or temperature or some aspect of health history. A feature is a function (possibly the identity function) of one or more measurements that is to be used to perform inference. Given a problem domain – such as network traffic – there are potentially many measurements one could take, some more useful for a given inference task than others. Given a set of measurements, there are many (infinitely many) features one could produce. Feature extraction will refer to both choosing the appropriate measurements to extract, and determining the features that are relevant to the problem. Feature selection is choosing the best – under some criterion – of features to use for inference.
Feature Extraction and Classification Techniques for Power Quality Disturbances in Distributed Generation: A Review
Published in IETE Journal of Research, 2023
Nivedita Singh, M.A. Ansari, Manoj Tripathy, Vivek Pratap Singh
This paper provides an exhaustive literature survey on the advancement of PQ analysis methodology, which is the central area of research into power systems. Researchers have tried to use a specific feature set to define the various PQ incidents. Consequently, there is a broad spectrum of literature in this area, the recent power requirements and the amount of micro-electronics accessible, and micro-electronic components/elements have raised concern about clean power and fault identification. That is why most of the works published have only been dated in the past few decades. There are many advanced signal processing for detection for PQ events/issues like Fourier Transform, Fast Fourier Transform, Wavelet Transform, S-Transform, and so on. After detection, data processing or feature extraction is performed; set of selected features is fed as input to classification purpose which is AI Technique, i.e. ANN, Fuzzy-Logic, Expert System, SVM, GA, and many more.
Wind speed forecasting using deep learning and preprocessing techniques
Published in International Journal of Green Energy, 2023
Management of input data is another way of improving the performance of the forecasting model. Feature selection and extraction are two ways to reduce the input data dimensionality. Feature selection selects the minimum subset of the input feature set while feature extraction is a mapping of the original set (Liu and Chen 2019a). Selecting the right set of features is extremely important since the selected feature set will be the only source of information for the learning algorithm. The aim is to avoid selecting too many or too few features. If too few features are selected, there is a chance that the set of features is low. If there are too many selected, then the effect of noise present in the data may minimize the information of the data set and might increase computational cost and overfitting (Senthil and Lopez 2015). The feature selection methods used in hybrid wind speed forecasting models are usually based on correlation, clustering, and information.
One-dimensional residual convolutional auto-encoder for fault detection in complex industrial processes
Published in International Journal of Production Research, 2022
Over the past two decades, machine learning methods have become popular in a variety of fields. For example, artificial neural network (Chen and Liao 2002; Yang 2015), support vector machine (SVM) (Yin and Hou 2016), support vector clustering (Yu 2013), Gaussian mixture model (GMM) (Yu and Qin 2009; Tong, Nael, El-Farra, and Palazoglu 2014), support vector data description (Jiang, Yan, and Lv 2014), and hidden Markov models (Yu 2010), have been applied in MSPC methods for process monitoring. However, it is difficult for these machine learning methods to learn effective features from signals directly for fault detection (Butte and Tang 2010). In general, a feature extraction step is generally performed to reduce the input dimension and extract features before the construction of these machine learning models.