機器學習 廣告欺詐檢測
Payment fraud has a long history and is the most common form of online fraud in the United States and the world. Recently, however, digital fraud has increased so much that it is difficult for many to distinguish the myth from reality.
付款欺詐行為源遠流長，是美國和全世界最常見的線上欺詐行為。 然而，近來，數字欺詐已大大增加，以至于許多人很難将神話與現實區分開。
Faced with the relentless threat of criminals, it has become indispensable to use state-of-the-art systems with learning capabilities, as companies strive to stay ahead. This reflects how organised crime and state-sponsored fraudsters are stepping up their fraud efforts.
面對犯罪分子的無情威脅，随着公司努力保持領先地位，使用具有學習能力的最新系統已變得不可或缺。 這反映了有組織犯罪和國家資助的欺詐者如何加強其欺詐工作。
The most common approaches to fighting online fraud include rules and prediction models that are no longer up to the complexity of today’s increasingly advanced online threats. The vast majority of new and emerging attacks in the field of digital fraud rely on machine learning and other automation techniques to commit fraud — an old approach to fraud prevention that cannot catch up.
打擊線上欺詐的最常見方法包括規則和預測模型，這些規則和預測模型已不再适應當今日益先進的線上威脅的複雜性。 數字欺詐領域中的絕大多數新出現的攻擊都依靠機器學習和其他自動化技術來進行欺詐，這是一種無法趕上的防止欺詐的舊方法。
Integrating AI-based platforms to detect online fraud into a high-risk game is a key part of the AI that today enables us to expand online fraud prevention. Digital businesses with a particular business model and their fraud analysts can take results from fraud analyses based on monitored and unattended machine learning to provide their business models with the information they need to detect and stop threats at an early stage. The results of unsupervised machine learning and supervised machine learning are characterised by the detection of anomalies in emerging data, and their integration into risk assessments is an important step towards enabling artificial intelligence to detect online crimes and enhance online prevention.
将基于AI的平台內建到高風險遊戲中以檢測線上欺詐是AI的關鍵部分，如今，它使我們能夠擴充線上欺詐預防。 具有特定業務模型的數字業務及其欺詐分析人員可以基于監視和無人值守的機器學習從欺詐分析中擷取結果，進而為業務模型提供在早期階段檢測和阻止威脅所需的資訊。 無監督機器學習和有監督機器學習的結果的特征在于檢測新興資料中的異常，将它們內建到風險評估中是邁向人工智能檢測線上犯罪并增強線上預防的重要一步。
Photo by Scott Graham on Unsplash Scott Graham在 Unsplash上 拍攝的照片
Financial institutions and other online traders are looking for innovative solutions that are well aligned with the system requirements. Machine Learning (ML) and Artificial Learning (AI) solutions promise to significantly reduce digital fraud. When consumers receive a call, SMS, email, or app message from their card issuer asking them to confirm a transaction or inform them of fraud on their cards, they can assume that a brilliant set of algorithms is just as excellent as customer service.
金融機構和其他線上交易者正在尋找與系統要求完全一緻的創新解決方案。 機器學習(ML)和人工學習(AI)解決方案有望顯着減少數字欺詐。 當消費者從發夾行收到呼叫，短信，電子郵件或應用消息，要求他們确認交易或通知他們卡上的欺詐行為時，他們可以認為一套出色的算法與客戶服務一樣出色。
Just after the use of artificial intelligence (AI) and machine learning have been invented and applied to payment fraud for the first time, the growing number of credit and debit card scams began to reduce. Combined with supervised and unsupervised methods, these models are able to learn and recognize new patterns that other approaches to fraud management may miss.
就在人們首次發明了使用人工智能(AI)和機器學習并将其應用于支付欺詐之後，越來越多的信用卡和借記卡騙局開始減少。 結合監督和非監督方法，這些模型能夠學習和識别其他欺詐管理方法可能會錯過的新模式。
Photo by Pascal Bernardon on Unsplash Pascal Bernardon在 Unsplash上的 照片
Other than AI technologies, Blockchain technologies also provide a higher level of security when reviewing financial transactions, and there is no doubt that today’s rule-based manual processes are largely incapable, making it harder than ever to evade fraud detections. As financial institutions use the latest fraud detection and prevention technologies, the combination of the two is becoming a viable alternative.
除了人工智能技術外，區塊鍊技術還可以在審查金融交易時提供更進階别的安全性，毫無疑問，當今基于規則的手動流程幾乎無能為力，這使得逃避欺詐檢測比以往更加困難。 随着金融機構使用最新的欺詐檢測和預防技術，将兩者結合起來已成為可行的選擇。
For instance, an AI-based technology provider company that specialises in selling fraud solutions to banks has helped Danske Bank reduce its false alarms by 60% and increase real fraud detection by 50%. This can be achieved not only by tracking down stolen identities but also by using artificial intelligence and blockchain technology.
例如，一家基于AI的技術提供商公司專門向銀行銷售欺詐解決方案，已幫助Danske Bank減少了60％的虛假警報，并将實際欺詐檢測提高了50％。 這不僅可以通過追蹤被盜身份來實作，還可以通過使用人工智能和區塊鍊技術來實作。
Other providers, such as healthcare providers and health insurers, are investigating AI to detect fraud. A recent survey by Optum found that 43% of healthcare industry leaders “strongly agree” that artificial intelligence will become an important tool to detect telemedicine fraud, waste, and abuse in reimbursement. Especially during the pandemic process, it is now even more critical.
其他提供商，例如醫療保健提供商和健康保險公司，正在調查AI以檢測欺詐行為。 Optum最近的一項調查發現，43％的醫療保健行業上司者“強烈同意”人工智能将成為檢測遠端醫療欺詐，浪費和報帳濫用的重要工具。 尤其是在大流行過程中，這一點現在變得更加關鍵。
The fight against online payment fraud in 2020 also highlights the security concerns associated with the rise of mobile apps and online platforms that make it easier to transfer money from one bank account to another, such as PayPal and MasterCard. Losses from digital money fraud are expected to increase by 130% by 2024, with emerging markets particularly vulnerable, according to Juniper. It is estimated that this will increase spending to $10 billion in 2024 — a 15% increase from 2020. Artificial intelligence and machine learning are a key component in the fight against online payment fraud, according to a recent International Monetary Fund (IMF) report.
2020年與線上支付欺詐的鬥争還凸顯了與移動應用程式和線上平台的興起相關的安全問題，移動應用程式和線上平台的興起使得将錢從一個銀行帳戶轉移到另一個帳戶(如PayPal和MasterCard)變得更加容易。 據瞻博網絡稱，到2024年，數字貨币欺詐造成的損失預計将增加130％，新興市場尤其脆弱。 據估計，到2024年，這将使支出增加到100億美元，比2020年增加15％。根據國際貨币基金組織(IMF)的最新報告，人工智能和機器學習是打擊線上支付欺詐的關鍵組成部分。
While technological advances and criminal machinations become more sophisticated, banks and financial institutions can harness the power of artificial intelligence to protect their businesses and improve the customer experience. The use of machine learning in fraud detection enables financial companies to identify real transactions and fraudulent transactions with greater accuracy in real-time. The potential for detecting and preventing fraud has had a significant impact on the growth of the financial services industry in recent years.
當技術進步和犯罪手段變得更加複雜時，銀行和金融機構可以利用人工智能的力量來保護其業務并改善客戶體驗。 在欺詐檢測中使用機器學習使金融公司能夠實時地，更準确地識别真實交易和欺詐交易。 近年來，發現和預防欺詐的潛力已對金融服務行業的增長産生了重大影響。
As fraud detection software based on artificial intelligence becomes more advanced and user-friendly, we will see this type of technology increasingly used in the area of digital fraud prevention and detection.
随着基于人工智能的欺詐檢測軟體變得更加先進和使用者友好，我們将看到這種類型的技術越來越多地用于數字欺詐預防和檢測領域。
Cited Sources
被引來源
-
https://www.forbes.com/sites/louiscolumbus/2019/08/01/ai-is-predicting-the-future-of-online-fraud-detection/
https://www.forbes.com/sites/louiscolumbus/2019/08/01/ai-is-predicting-the-future-of-online-fraud-detection/
-
https://pegus.digital/how-ai-can-help-combat-fraud/
https://pegus.digital/how-ai-can-help-combat-fraud/
-
https://www.fico.com/blogs/5-keys-using-ai-and-machine-learning-fraud-detection
https://www.fico.com/blogs/5-keys-using-ai-and-machine-learning-fraud-detection
-
https://www.bankingexchange.com/big-data/item/8140-machine-learning-and-ai-crucial-to-fighting-fraud-research-shows
https://www.bankingexchange.com/big-data/item/8140-machine-learning-and-ai-crucial-to-fighting-fraud-research-shows
-
https://dzone.com/articles/how-ai-companies-are-gearing-up-to-mitigate-digita
https://dzone.com/articles/how-ai-companies-are-gearing-up-to-mitigate-digita
-
https://www.netguru.com/blog/fraud-detection-with-machine-learning-banking
https://www.netguru.com/blog/fraud-detection-with-machine-learning-banking
-
https://www.zmescience.com/science/ai-fraud-detection-0942323/
https://www.zmescience.com/science/ai-fraud-detection-0942323/
翻譯自: https://medium.com/swlh/machine-learning-based-digital-fraud-detection-bad492232eb6
機器學習 廣告欺詐檢測