Agricultural Bank Data Mining Competition: Fourth Paradigm AutoML Wins Championship

2019-11-28

Recently, the award ceremony of the "Athena Cup" data mining competition, jointly organized by the Agricultural Bank of China and the "Financial Electronicization" magazine, was held in Beijing. The team from Fourth Paradigm (AutoXing) won the first place in the competition and was awarded by Zhang Qingsong, the President of the Agricultural Bank of China. The "Athena Cup" data mining competition has reached a high level in terms of participation breadth and depth, attracting 361 teams from over 70 financial institutions, technology companies, and universities, with a total of 870 participants. The competition aims to solve practical problems faced by financial enterprises in their business development through methods such as machine learning. The competition theme is set as loan risk prediction, where each participating team needs to establish an accurate risk control model and predict whether users will default on their loans based on loan data information. The final evaluation of the participating teams focused on their understanding of business scenarios and data, feature extraction, and high-dimensional algorithm modeling capabilities. The Fourth Paradigm team utilized the Fourth Paradigm AutoML technology to construct features from different dimensions, including user basic attributes, credit card billing records, browsing behavior, and bank transactions. They performed feature selection from multiple perspectives to reduce noise and improve model stability. Through model fusion, they further enhanced the model's effectiveness, ultimately winning the championship in this competition. AutoML (Automatic Machine Learning) is a low-threshold AI application product developed by Fourth Paradigm, aiming to automate the modeling and parameter tuning process, making the entire machine learning process more automated and reducing reliance on AI scientists. Since 2015, Fourth Paradigm has focused on the forefront innovation and practical application of AutoML technology, validating its capabilities and commercial value in real scenarios. Currently, many of Fourth Paradigm's clients have independently built AI applications based on the AutoML product in various business scenarios. In this competition, a joint-stock commercial bank also achieved outstanding performance using Fourth Paradigm's AutoML technology. Fourth Paradigm has always been committed to integrating artificial intelligence technology with the production and operation of various industries. Dai Wenyuan, the founder and CEO of Fourth Paradigm, introduced, "Fourth Paradigm lowers the threshold for AI implementation and accelerates AI innovation through its platform, helping enterprises achieve AI transformation. Currently, we have empowered tens of thousands of scenarios in the fields of finance, energy, retail, government affairs, healthcare, security, media, and more. With the help of AI, we will discover that various industries are entering the scientific era, where machines discover patterns and predict demands from data, enabling more accurate customized services and maximizing resource allocation, significantly improving operational precision." Zhang Yongfu, the Secretary of the Party Committee and General Manager of the China Financial Electronicization Company of the People's Bank of China, expressed that this competition brought together outstanding data mining participants in the industry, showcasing their unique understanding of the financial industry's business and technological advantages. A number of excellent model achievements stood out, providing many outstanding examples for the technological leadership of the financial industry.

RECENT AI NEWS

RECENT AI TOOLS