diff --git a/docs/gallery/articles/fraud_detection/images/analysis_page.png b/docs/gallery/articles/fraud_detection/images/analysis_page.png new file mode 100644 index 000000000..09ae3e7cc Binary files /dev/null and b/docs/gallery/articles/fraud_detection/images/analysis_page.png differ diff --git a/docs/gallery/articles/fraud_detection/images/fraud_explanation.png b/docs/gallery/articles/fraud_detection/images/fraud_explanation.png deleted file mode 100644 index c96e922e9..000000000 Binary files a/docs/gallery/articles/fraud_detection/images/fraud_explanation.png and /dev/null differ diff --git a/docs/gallery/articles/fraud_detection/images/fraud_threshold.png b/docs/gallery/articles/fraud_detection/images/fraud_threshold.png deleted file mode 100644 index 4ea3147a6..000000000 Binary files a/docs/gallery/articles/fraud_detection/images/fraud_threshold.png and /dev/null differ diff --git a/docs/gallery/articles/fraud_detection/images/fraud_transactions.png b/docs/gallery/articles/fraud_detection/images/fraud_transactions.png deleted file mode 100644 index 8f666d2a2..000000000 Binary files a/docs/gallery/articles/fraud_detection/images/fraud_transactions.png and /dev/null differ diff --git a/docs/gallery/articles/fraud_detection/images/login_page.png b/docs/gallery/articles/fraud_detection/images/login_page.png new file mode 100644 index 000000000..be1387b1d Binary files /dev/null and b/docs/gallery/articles/fraud_detection/images/login_page.png differ diff --git a/docs/gallery/articles/fraud_detection/images/threshold_page.png b/docs/gallery/articles/fraud_detection/images/threshold_page.png new file mode 100644 index 000000000..069827e2a Binary files /dev/null and b/docs/gallery/articles/fraud_detection/images/threshold_page.png differ diff --git a/docs/gallery/articles/fraud_detection/images/transactions_page.png b/docs/gallery/articles/fraud_detection/images/transactions_page.png new file mode 100644 index 000000000..da9a8d978 Binary files /dev/null and b/docs/gallery/articles/fraud_detection/images/transactions_page.png differ diff --git a/docs/gallery/articles/fraud_detection/images/user_page.png b/docs/gallery/articles/fraud_detection/images/user_page.png new file mode 100644 index 000000000..805dfece2 Binary files /dev/null and b/docs/gallery/articles/fraud_detection/images/user_page.png differ diff --git a/docs/gallery/articles/fraud_detection/index.md b/docs/gallery/articles/fraud_detection/index.md index 9f02acfca..bfd84f506 100644 --- a/docs/gallery/articles/fraud_detection/index.md +++ b/docs/gallery/articles/fraud_detection/index.md @@ -4,38 +4,64 @@ category: finance data-keywords: dashboard vizelement layout chart ai multi-page classification enterprise short-description: A Taipy Application that analyzes credit card transactions to detect fraud. order: 20 -img: fraud_detection/images/fraud_threshold.png +img: fraud_detection/images/transactions_page.png hide: - toc --- -A Taipy Application that analyzes credit card transactions to detect fraud. + +A Taipy Application to analyze credit card transactions, detect fraud, and collaborate with other +application users. !!! note "Taipy Enterprise edition" Taipy provides robust, business-focused applications tailored for enterprise environments. To maintain standards of security and customization, these applications are proprietary like this - application. If you’re looking for solutions that are immediately deployable and customizable to + application. If you're looking for solutions that are immediately deployable and customizable to your business needs, we invite you to try them out and contact us for more detailed information. - [Try it live](https://fraud-detection.taipy.cloud/Transactions){: .tp-btn target='blank' } + [Try it live](https://fraud-detection.taipy.cloud/Login){: .tp-btn target='blank' } [Contact us](https://taipy.io/book-a-call){: .tp-btn .tp-btn--accent target='blank' } # Understanding the Application -This application shows a list of credit card transactions. The user can select a date range to -predict fraud. The application will then use an XGB model to mark potentially fraudulent -transactions in red or yellow. +This application displays a list of credit card transactions. A model estimates whether a +transaction is fraudulent; this task can be automatically handled by a pipeline. However, +some transactions may require further human review. + +![Transactions](images/transactions_page.png){width=90% : .tp-image-border } + +Within this page, you can access various analyses and data visualizations: + +- List of transactions +- Client information +- Fraud details + +This demo includes user management and collaboration features. You need to select one of the +available users to access the application. + +![Users](images/login_page.png){width=90% : .tp-image-border } + +After logging in, you can navigate to your user page to view the transactions assigned to you +for investigation. You can see both your past transactions and those requiring your attention. +Clicking on a transaction in the table will select it and navigate you to the Analysis page. + +This page also includes a newsfeed showing the application or other users' activities. + +![User Page](images/user_page.png){width=90% : .tp-image-border } -![List of Transactions Page](images/fraud_transactions.png){width=90% : .tp-image-border } +The Analysis page presents several pieces of information. The left section explains the model's +results (providing explanations on the model output), the middle section displays details about +the transaction, and the right section shows information about the client. Here, you can verify +the client's identity using a deep learning model. -The user can select a transaction to see an explanation of the model's prediction, as well as the client's -other transactions. +You can decide whether the transaction is fraudulent or not. If you are unsure, you can share the +transaction with someone else for further review. -![Prediction Explanation Page](images/fraud_explanation.png){width=90% : .tp-image-border } +![Analysis](images/analysis_page.png){width=90% : .tp-image-border } -The user can also choose the threshold of the model. The threshold is the model output -above which a transaction is considered fraudulent. The user can select the model according -to the displayed confusion matrix and by looking at False Positive and False Negative transactions. +For educational purposes, you can adjust the model's threshold— the output value above which a +transaction is considered fraudulent. You can select the threshold by examining the displayed +confusion matrix and reviewing false positive and false negative transactions. -![Threshold Selection Page](images/fraud_threshold.png){width=90% : .tp-image-border } +![Threshold Selection Page](images/threshold_page.png){width=90% : .tp-image-border } \ No newline at end of file