The study is combined with the text mining and classification method to analyze the tumor-related tweets on Twitter in order to comprehend the public opinion focus on the social media. The contribution of the study in...
详细信息
The study is combined with the text mining and classification method to analyze the tumor-related tweets on Twitter in order to comprehend the public opinion focus on the social media. The contribution of the study included the following two points. First, the study used the text mining method to explore the content of tumor-related tweets and found the keywords according to their frequencies. Second, the study applied the classification analysis with four different algorithms to explore the relationship and the importance of the keywords. The study also found the random forests model with the best performance in four models and concluded that the Twitter users focused more on surgery issues, psychological issues and the technical issues on the topic of tumor.
The use of polymer informatics is an appropriate solution to overcome the problems of optimizing the synthesis conditions of polymers, which has attracted the attention of many researchers. The aim of this study is to...
详细信息
In this paper, the internal energy consumption on the Internet of Things (IoT) has been studied. The purpose of this paper is to predict the factors affecting energy consumption in buildings by considering machine lea...
详细信息
In this paper, the internal energy consumption on the Internet of Things (IoT) has been studied. The purpose of this paper is to predict the factors affecting energy consumption in buildings by considering machine learning algorithms such as k-nearest neighbors (KNN), adaboost, random forest and neural network. These algorithms are implemented in the Orange tool. Also, the univariate regression algorithm is used to select the best feature. This algorithm determines the most important factors affecting energy consumption and their impact. Then, with the help of Gephi tool, these data are simulated in the IoT environment as a complex network. The simulated network in the Internet of things is also a small world network. This network shows the relationships between the features. The results of this paper show that the overall height, roof area, surface area and relative compaction have the greatest impact on the energy consumption of buildings. It can be seen that the predicted error percentages for these data for cooling loads and heating loads are 0.911 and 0.292, respectively. It should be noted that the best algorithm for cooling load and heating load is ada boost algorithm.
Fault diagnosis by machine learning techniques is especially crucial to maintain the stability of the water processing flow in wastewater treatment plants(WWTPs). A key factor influencing the accuracy of fault diagnos...
详细信息
Fault diagnosis by machine learning techniques is especially crucial to maintain the stability of the water processing flow in wastewater treatment plants(WWTPs). A key factor influencing the accuracy of fault diagnosis lies in the imbalance data distribution between majority classes(i.e., normal situations) and minority classes(i.e., faulty situations), which may cause misjudgments of faults and lead to failure in practical use. This study proposes an ensemble classification method called ada WELM ensemble algorithm for fault diagnosis in wastewater treatment which builds up individual classifiers by using weighted extreme learning machine(WELM) and combines them with adaboost ensemble algorithm. The weight matrix in WELM can be updated adaptively along with the iteration of adaboost learning. The simulations based on 25 benchmark datasets from KEEL repository are arranged at first and the results show the proposed classifier outperforms the other comparative classifiers for most of the evaluated datasets. And then, a practical fault diagnosis model in wastewater treatment plant based on ada WELM ensemble algorithm is built and the simulations verifies that, comparing with several classifiers, this fault diagnosis model can efficiently improve the performance of WWTP, especially improve the accuracy of faulty class and G-mean value. So we think the proposed ada WELM ensemble method could be a promising solution for imbalanced data classification in *** experiments, we find that the performance of G-mean value and accuracy of minority class in a case study of WWTPs are improved by comparing with several other related classification algorithms.
Most of the state-of-the-art link prediction algorithms for social networks have the disadvantage of low recall ratio, due to the fact that the adjacency matrices are usually extremely sparse in practical situations, ...
详细信息
Most of the state-of-the-art link prediction algorithms for social networks have the disadvantage of low recall ratio, due to the fact that the adjacency matrices are usually extremely sparse in practical situations, which make it harder to identify link patterns even when they are actually present in the networks. In this article, we propose an ensemble link prediction framework to address this problem. By carefully exploring the characteristics of typical approaches that have been proposed in previous studies, we identify six approaches that are complementary to each other in terms of the true positive and false positive rates. And then we propose an ensemble approach, called ada Pred, for predicting implicit links in social networks, based on fusion of the six individual weak learners. Experimental results on real datasets show that the predictive accuracy and the recall ratio of the proposed ada Pred algorithm consistently outperform other methods reported in the literature.
In order to solve the problems of detecting and tracking human face in multiple-view and under complicated background by using camera in real-time, this paper puts forward a new way that combines the multi-view face d...
详细信息
ISBN:
(纸本)9781479921867
In order to solve the problems of detecting and tracking human face in multiple-view and under complicated background by using camera in real-time, this paper puts forward a new way that combines the multi-view face detection algorithm and face tracking algorithm to control a rotatable camera. The built-in system is able to track a human face through a camera, in which adaboostalgorithm and Continuously adaptive Mean Shift (Cam Shift) algorithm are adopted to detect and track the face, respectively. For reducing the face detection and tracking time, the preprocessing methods of skin region detection and image sampling are used. The latter one showed improved performance in face detection and tracking far better than the former in real-time. The experimental results show that the system can keep the tracked human face in the middle of visual field of the camera approximately with high accuracy and low cost.
This paper introduces a method for vehicle detection based on a machine learning algorithm, which is capable of rapidly detecting a vehicle and achieving high detection rates. Firstly, a Haar-like feature is used to r...
详细信息
ISBN:
(纸本)9781784660529
This paper introduces a method for vehicle detection based on a machine learning algorithm, which is capable of rapidly detecting a vehicle and achieving high detection rates. Firstly, a Haar-like feature is used to represent the appearance of the vehicle, then a learning algorithm, based on adaboost, is used to train the strong classifier, and lastly a method for combining strong classifiers in a cascade is proposed, which allows background regions of the image to be quickly removed. The experimental results show that our classifier can achieve good performance of vehicle detection;its detection rate is more than 97% and its false alarm rate is below 0.2%.
暂无评论