KEYWORDS: Tunable filters, Social networks, Data modeling, Performance modeling, Matrices, Linear filtering, Digital filtering, Detection theory, Neural networks, Data mining
Graph anomaly detection in graph data has received significant attention due to its practical significance in various vital applications such as network security, finance, and social networks. The current mainstream approach for attribute graph anomaly detection is based on contrastive learning using graph neural networks, which only consider homogeneous low-frequency signals. However, in attribute networks, normal and anomalous nodes exhibit different frequency patterns. This motivates the proposal of a graph anomaly detection framework based on multi-frequency reconstruction to capture the signal patterns of anomaly. Specifically, our method constructs multiple filters based on target nodes and utilizes two modules, namely, low-frequency reconstruction and contrastive learning, for anomaly detection. The generative low-frequency reconstruction module enables us to capture anomalies in the high-frequency attribute space, while the contrastive learning module leverages richer structural information from multiple subgraphs to capture anomalies in the structural and mixed spaces. We conducted extensive experiments on five publicly available datasets, demonstrating that our method significantly outperforms state-of-the-art approaches.
KEYWORDS: Data modeling, Performance modeling, Matrices, Feature extraction, Design and modelling, Proteins, Neural networks, Data processing, Data mining
Autoencoders, as a type of generative self-supervised learning, have received increasing attention in information processing and data mining in recent years. However, existing autoencoders usually generate graph data conforming to the feature distribution from only one aspect of reconstructing edge or node features, which allows the models to extract only a single level of information, limiting their application in real-world applications. In this paper, we propose DummyMAE, a generative self-supervised learning framework that synchronously generates edge and node features. In general, it losslessly converts vertex graphs into corresponding line graphs by introducing edge-to-vertex transformations. The vertex graph provides the model with information on node features, and the line graph provides the model with the ability to capture information on the graph structure, which complements each other. The task of simultaneously reconstructing edges and features is achieved in this way. The task of graph classification serves as a pivotal component within the realm of graph learning, we have conducted sufficient experiments on four widely used graph classification datasets, and the results show that DummyMAE outperforms the current state-of-the-art baselines for the graph classification task.
Food adulteration driven by economic interests is an important cause of food safety. Camel milk is widely sought for its high nutritional and medicinal value; some businesses adulterate it for profiteering due to its low yield and high price. Traditional adulteration detection methods rely on supervised learning, which is limited by the data of unknown categories in practical application scenarios, and it is difficult to solve the adulteration problem of category imbalance. For the above scenario, this paper proposes a camel milk adulteration detection framework FIAD based on an unsupervised anomaly detection algorithm, which starts from the perspective of anomaly detection and automatically captures and isolates anomalous features in the data through a tree algorithm without manual labeling of data, directly targeting the adulteration identification problem of category imbalance. We tested the discrimination performance of FIAD in a batch of category-imbalanced camel milk adulteration datasets. At 10% and 20% category imbalance, FIAD achieved AUCs of 0.943 and 0.959 and Recall of 0.915 and 0.949, while occupying less memory, better than eight baseline models. The results show that FIAD has excellent comprehensive identification performance and provides a low-cost and high-efficiency identification method for camel milk adulteration identification.
Cervical cancer is one of the most common female malignant tumors in the world. In recent years, the incidence of cervical cancer has tended to be younger, which has attracted great attention from all countries in the world. Early and accurate diagnosis of cervical cancer is of great significance to patients. At present, the common diagnostic methods of cervical cancer in China include cytological screening and HPV detection, but these methods are generally greatly affected by doctors' subjective factors and cannot fully meet the domestic clinical needs, so a rapid and accurate diagnosis of cervical cancer is of great value for exploration and research. In this paper, serum infrared spectroscopy technology combined with machine learning was used to diagnose and classify cervical cancer patients. Firstly, the spectral data were preprocessed by smoothing and normalization, and principal component analysis (PCA) was used to reduce dimension. The obtained data were imported into Support Vector Machines with Particle Swarm Optimization (PSO-SVM), Random Forest (RF) and K-Nearest Neighbor (KNN) models for classification, and ten-fold cross-validation was used to verify the performance of the model. Finally, the established models are compared.
Cervical cancer is one of the major gynecological malignancies that seriously endanger women's health. Patients with early symptoms are not obvious and prone to metastasis and recurrence, leading to poor prognosis of patients with cervical cancer. At present, cytological screening and HPV detection are the main diagnostic methods of cervical cancer in China, but both of them are greatly influenced by doctors' subjective factors, with low specificity and high rate of missed diagnosis. Therefore, a rapid and effective diagnostic method is needed to be explored. In this paper, the serum samples of patients with cervical cancer were taken as the research object, and the experimental serum samples were analyzed by infrared spectroscopy, which provided a clinical basis for the identification and classification of patients with cervical cancer by infrared spectroscopy. In this study, infrared spectral signals of serum of patients with cervical cancer were collected, and spectral signals were analyzed and preprocessed. Partial least squares regression (PLS) was used to select spectral signal features. Then, an Xgboost ensemble learning model is established using GBtree, GBlinear and Dart as the base classifier, and the performance of the model is evaluated by using the ten-dot cross-validation. Finally, the established models are compared.
In recent years, water quality testing has become an increasingly important topic. Compared with some common water quality identification methods, this study proposes a new method for identifying water samples in UV-visible spectroscopy. In this study, the UV-visible spectra of water samples from two different regions of tianchi and shuimogou in Urumqi were measured, and the pattern recognition algorithm was used to identify the two types of water samples. The acquired UV-visible spectra of water samples were extracted from 80 original high-dimensional spectral data by Partial Least Squares Regression (PLS), and the extracted features were modeled and classified by Support Vector Machine (SVM) classifier. The parameters C and g are optimized by Grid Searching (GS). The classification accuracy of the tianchi water sample and the water mill ditch water sample was 100%. The results of this study illustrate the great potential for rapid detection of water samples using UV-visible spectroscopy in the future.
Arsenic (As) is a trace element exist in the environment, and it is one of the common poisonous elements in water, excessive intake of arsenic can cause great damage to human body. At present, mainly used laboratory detection methods of arsenic such as electrochemical method, ion chromatography, atomic absorption spectroscopy and so on, can detective arsenic, but these techniques have some problems such as low sensitivity, intractable operation and expensive. Based on the specific molecules of arsenic, we tested a new rapid detection method of arsenic solution, we prepared surface-enhanced Raman enhanced scattering substrate (SERS substrate) to complete the detection of arsenic solution. Through linear discriminant analysis, the result show that Raman spectrum has high specificity and sensitivity. The study indicated the feasibility of using SERS substrate to conduct Raman spectrum detection on arsenic, which was of great significance for the detection of arsenic in aqueous solution.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.