China
Africa
Explore chapters and articles related to this topic
Introduction
Published in Junzhi Yu, Xingyu Chen, Shihan Kong, Visual Perception and Control of Underwater Robots, 2021
Junzhi Yu, Xingyu Chen, Shihan Kong
On the basis of YOLOv2, Redmon et al. proposed YOLOv3 [31] and introduced a multi-scale prediction mechanism (introduced in Section 2.3.2.2). In addition, logistic regression, instead of traditional softmax, was used for predicting detection confidence. Since softmax can be essentially replaced by multiple independent logistic regression classifiers, where inter-category interference is relatively small, this method is more suitable for multi-label classification.
A linear space adjustment by mapping data into an intermediate space and keeping low level data structures
Published in Journal of Experimental & Theoretical Artificial Intelligence, 2021
Weiqing Fu, Hamid Parvin, Mohammad Reza Mahmoudi, Bui Anh Tuan, Kim-Hung Pho
Multi-class learning is a learning model in which the training task is not different from test task and there are more than two classes in the problem (Rajasekar and Er (2016)). It can be solved by One-Vs-One approach or One-Vs-All approach. It can be aggregated by the space adjustment if the given data set has more than 2 classes. Multi-instance learning (Zhu et al. (2017)) is a learning model in which bags of samples with label, instead of labelled samples, emerge. It is massively related to multi-label classification which is another topic related to the paper. In multi-label classification, each training sample may have more than one label. It differs from the space adjustment because, in multi-label classification, the data points in the training-data space and the data points in test-data space have the same distribution (Rajasekar et al. (2017)).
Multi-label classification algorithms for composite materials under infrared thermography testing
Published in Quantitative InfraRed Thermography Journal, 2022
Muflih Alhammad, Nicolas Peter Avdelidis, Clemente Ibarra Castanedo, Xavier Maldague, Argyrios Zolotas, Ebubekir Torbali, Marc Genest
Unlike binary or multiclass classification problems, multi-label classification is a classification problem, where each instance is associated with multiple target labels instead of only one. In other words, the task in this type is to predict the label-sets of unseen instances, instead of a single label. For multi-label classification, what so called problem transformation method and algorithm adaptation method are used. In the first approach, classification algorithms attempt to convert the original multi-labelled data into binary or multiclass format. The resulting predictions are then combined to form the output label-sets. In algorithm adaptation type, multiclass models such as Random Forest (RF) algorithm, which has been proposed for this investigation, are first modified to suit the nature of the multi-labelled data and are then applied directly to the problem [42, 43]. RF is a supervised ensemble of decision trees technique that uses a bagging method to construct a large number of relatively uncorrelated tress during the training process for the final prediction. The advantage of having many uncorrelated models is that the trees will protect each other from their individual errors. This structure, in turn, improves the accuracy and stability of the model by reducing the variance without increasing the bias and/or the noise. Each tree model is made up of a series of nodes and branches and splits the data at the root into subsets based on different features. At the end of the process, each individual tree will provide its prediction and classes with the most votes will represent the model prediction. It is worth pointing out that RF can also be used to identify the most significant variables in a given dataset from many input features.
Assessing injury severity of secondary incidents using support vector machines
Published in Journal of Transportation Safety & Security, 2022
Jing Li, Jingqiu Guo, Jasper S. Wijnands, Rongjie Yu, Chengcheng Xu, Mark Stevenson
The OL model has been widely used in fitting structure for an ordinal response. Injury severity of secondary incidents can be defined as ordinal variables that are classified into three levels: 1-no injury and only property damage; 2-minor injury; and 3-incapacitating and fatal injuries. Prediction rates of various models were assessed using accuracy, weighted precision and weighted recall. For multi-label classification, precision and recall of each label can be calculated.