Artificial Intelligence Lab
PayEgis Artificial Intelligence Laboratory is committed to combining cutting-edge technology with anti-fraud business, taking customer demand as the driving force and technological innovation as the core to provide advanced and intelligent risk control solutions for multiple industries.PayEgis artificial intelligence laboratory has promoted the research and development of a variety of anti-fraud products. Through distributed parallel computing, storage management, real-time retrieval and other technologies, it has achieved efficient monitoring analysis and disaster recovery backup capabilities. Based on the analysis capabilities of massive business data, it has identified fraud risks and guaranteed business security. At present, it has provided professional risk control solutions and efficient service support for Internet finance, power system, e-commerce, advertising and marketing and many other industries.
Knowledge map has become a research hotspot in the field of artificial intelligence in recent years due to its efficient retrieval ability and powerful visual relationship combing ability. PayEgis AI Lab is dedicated in researching knowledge graph technologies, exploring its application in risk management field, including the research of Neo4j、ArangeDB、JanusGraph and other mainstream graph database and the research of related analyze models like Community Detection, degree analysis, consistency check and other commonly used graph analysis models.
Among them, the community found that through the intelligent analysis of complex networks, it could identify the clustering risks under the huge social networks. With the maturity of knowledge mapping technology and related algorithm technology, it has become a hot research topic in the field of risk control in recent years. Artificial intelligence laboratory focuses on the exploration and application of the mainstream community discovery algorithm model. Through the integration of multiple models such as clustering, tag propagation, dyeing and correlation analysis, sophisticated risk control means such as fraud risk identification in complex social networks, black production link combing and device group control perception can be realized.
PayEgis artificial intelligence laboratory strives to use cutting-edge machine learning modeling technology to efficiently solve the customer's risk identification and risk prevention and control needs. K-means, DBSCAN, Louvian and other clustering models were used to identify potential risks in mass data. Logistic regression, Xgboost, LightGBM and other advanced classification integration models were used to evaluate the potential fraud risk of users. Finally, intelligent Suggestions are proposed for terminal risk control rules through SVM, decision tree, GBDT and other decision models. Ai lab is committed to providing advanced technical and theoretical support for anti-fraud product lines through cutting-edge exploration of machine learning.
The field of deep learning in machine learning has always been the key research direction of PayEgis artificial intelligence laboratory. As a hot research field under machine learning, deep learning and neural network-related technologies can meet the demand for risk control with large data volume and high computational efficiency, and also provide more efficient technical support for risk control means such as face recognition and in vivo detection. Based on mainstream deep learning frameworks such as Tensorflow and Pytorch, PayEgis artificial intelligence laboratory is committed to exploring front-end areas such as computer vision, natural language processing, speech recognition, and deep neural network, and provides all-round technical support for PayEgis anti-fraud products.
Products and Applications
PayEgis Graph Engine,based on Janus graph database, provides a visual map analysis and display platform for entity relationship association analysis. Through the establishment of complex relationship network, the product visualizes the relationship between entities, and is equipped with graph engine, relationship reasoning engine and other in-depth analysis tools to provide customers with necessary data analysis support. The relationship map provides advanced screening and early warning scheme for accurately exposing fraud ring, nest case, intermediary fraud, money laundering and other complex fraud means by combing, analyzing and graphically displaying the context of entity relationship, combining with intelligent analysis functions such as social network discovery and inconsistency test.
PayEgis modeling workbench,It integrates a variety of visual and interactive modeling workbench of machine learning models, and provides customers with simple and easy-to-use modeling solutions. It covers all the common algorithms in the field of machine learning, such as clustering, classification, outlier detection, unsupervised training, etc., and realizes the configurable model training through the built-in model framework, covering all the modeling processes such as data processing, feature processing, algorithm optimization, model deployment and iteration.
Identity authentication of PayEgis,Through computer vision algorithm based on deep neural network, accurate face comparison and in vivo detection function are provided. Combining with various authentication methods such as fingerprint probe technology and biological fingerprint identification, an accurate and efficient identity authentication channel is provided.