In October 1950, Alan Turing published his paper computational machinery and intelligence, put forward the famous “Turing test”, and set off the first wave of artificial intelligence. At the summer academic seminar held by Dartmouth University in 1956, J McCarthy, m Minsky and Shanno formally put forward the term “artificial intelligence (AI)”. However, due to the limitations of computer computing power and semiotic theory, the heat of artificial intelligence subsided rapidly. In the 1980s, artificial intelligence rose again. The breakthrough came from abandoning symbolism and changing to statistical thinking.
In 1991, IBM dark blue defeated chess player Kasparov to set off the second wave of artificial intelligence, but artificial intelligence at that time was limited by the amount of data and test environment and did not have practical value. The third wave of artificial intelligence originates from the concept of deep learning proposed by Hinton et al. In 2006.
At this time, the development of the Internet industry has formed a large amount of data, and the innovation of GPU has greatly improved the computing power of computers. With the maturity of these two conditions, artificial intelligence has achieved a leap forward development. Taking the victory of Google alphago over go player Li Shishi in 2016 as a landmark event, the concept of AI has been deeply rooted in the hearts of the people, theory has become practice, and scientific research has moved to the market.
Similar to the three waves of artificial intelligence, the development of fintech can also be divided into three stages, corresponding to different technology enabled financial industries:
1.0 e-finance, that is, in the period of “it + finance”, the development of information system promotes the electronization and automation of financial business, improves the service capacity of outlets, and reduces the transaction cost of deposit, loan and remittance;
2 online finance, namely Internet plus finance, has greatly extended the scope and depth of financial services with the generation of massive user data. For example, Alipay, which was born in 2003, promoted mobile payment business. In addition, the birth of information interaction channels has made financial services change, resulting in online banking and mobile phone banking.
3.0 intelligent finance, i.e. “Ai + finance”, with the maturity of new technologies (mainly ABCD, i.e. artificial intelligence (AI), blockchain, cloud computing and big data), further promote the intellectualization of the financial industry, reduce information asymmetry, reduce transaction costs, improve risk control level, and fully meet the personalized needs of customers.
Fintech mainly refers to the integration of big data, artificial intelligence, blockchain, cloud computing and other technologies with the financial industry. The application of fintech by central banks around the world is mainly a means to strengthen supervision and compliance through the above several core technologies. Rapidly improve the service efficiency, user experience and precise supervision of enterprises in the financial industry through emerging technologies.
In a broad sense, the development of artificial intelligence needs to obtain enough data for AI to learn based on sufficient computing power, and blockchain mainly plays the role of distributed information encryption. From a historical point of view, human beings have gradually changed from experimental science (drilling wood for fire, friction electrification, etc.) to theoretical science (Interpretation of natural phenomena), then to Computational Science (simulated phenomena), and finally to data Science (Law of massive data discovery).
At present, there are three cloud enterprises (Microsoft, Amazon and Google) in the field of big data storage and cloud computing. At the same time, with the rapid development of data storage, transmission and cloud computing, the development of AI can also be rapidly improved with the massive distributed encrypted information.
Machine learning machine learning is the foundation and core of artificial intelligence. It deduces the future variable events through probability, complex mathematical model, human behavior, psychology, combined with long-term and short-term memory neural network, convolutional neural network, deep confidence network, trestle self coding neural network and so on. As a complex machine learning algorithm, deep learning has the ability of analysis and learning to recognize text, image, sound and other data. It has good applicability to all kinds of financial data. It is widely used in stock market prediction, risk assessment, event early warning and so on.
Natural language process (NLP) NLP is a technology integrating linguistics, psychology, computer science, neuroscience and other fields. This technology has several important development significance. By better analyzing the relevance in language, it can improve the effectiveness of information extraction and self generate language information. NLP can automatically read and generate financial statements and research reports. Another important application is intelligent customer service, which will greatly improve the efficiency of financial institutions and reduce repetitive confirmation.
The application of knowledge map technology is mainly in the fields of customer marketing, anti fraud and anti money laundering. In the process of customer marketing, link multiple data sources in advance, predict the portrait and description of user groups in advance, and improve the accuracy of marketing. In anti fraud, identify fraud in advance to better verify the consistency of information. In anti money laundering, by screening suspicious transactions and suspicious cases in advance, the operation efficiency of financial institutions is greatly improved and human resources are saved. Finally, in investment research, it can also improve investment accuracy and information reading efficiency based on a large number of related knowledge.
Computer vision computer vision technology is mainly used in authentication, mobile payment, user security and other fields. In the aspect of identity verification, the intelligent device captures the face and extracts the certificate information, extracts the face features and face key points, and confirms the identity security. At the mobile payment end, more secure and fast payment can be made through face recognition and specific characteristic behavior . Banks will be able to avoid the potential risks in large transactions through computer vision, and provide more efficient and safe services by analyzing transaction scenarios.
Application of artificial intelligence in financial scene
At present, one of the most extensive applications of artificial intelligence is customer service. The most common methods include customer communication channels, services adjusted in real time according to customer needs, and targeted risk exposure analysis. The earliest application of artificial intelligence in customer service stems from creating more convenient payment methods. In order to simplify the cumbersome processes such as repeated authentication and bank account authentication in the process of payment and transfer, financial institutions have designed authentication tools such as image CV and biometrics to enable investors to obtain a better user experience.
For financial institutions, the application of AI can greatly reduce labor service costs, improve process efficiency and improve the interaction mode with investors. The company continuously collects data related to products, services and processes through business accumulation, and uses AI to process and classify the data, so as to answer different types of investment problems.
The main objectives of the operation are to expand users, improve user activity, improve product experience and find a profit model. The operation of the financial industry generally includes user operation, channel operation, product operation, activity operation, data operation and brand operation. The particularity of the financial industry also determines the need for a deeper understanding of the market and risk in the process of obtaining customers, serving and maintaining. 2.2.1 stress testing after the 2008 financial crisis, Basel III requires banks to conduct stress testing within each specified cycle, but this is a new challenge for banks. The requirements of stress test for VaR, CVaR, core capital adequacy ratio and liquidity adequacy ratio have greatly increased the amount of data and workload. As shown in Figure 5, the impact of the epidemic on large commercial banks includes multi-dimensional data.