Ce diaporama a bien été signalé.
Nous utilisons votre profil LinkedIn et vos données d’activité pour vous proposer des publicités personnalisées et pertinentes. Vous pouvez changer vos préférences de publicités à tout moment.

AI Fraud Detection - by Edyoda

293 vues

Publié le

"Development to deployment of Fraud Detection Infrastructure",
Speaker: Awantik Das: Co-Founder - edyoda (https://www.linkedin.com/in/awantik/)

Talk high level summary: Credit card fraud detection is heavily used application in banks & financial institutions. In this workshop, the entire infrastructure of building complete system of credit card fraud detection will be discussed. It will include Kafka for getting transaction records, building & deploying machine learning models for this application. Finally, storing the predictions in cassandra will be explored. It will be a complete hands-on session and would also include bit of introduction of Cassandra and Kafka.

Presented as part of the AI & ML meetup in NVidia: https://www.meetup.com/Bangalore-AI-ML-Meetup/events/262438475/

Publié dans : Logiciels
  • Soyez le premier à commenter

  • Soyez le premier à aimer ceci

AI Fraud Detection - by Edyoda

  1. 1. Credit Card Fraud Detection By Awantik Das, Edyoda
  2. 2. What is Credit Card Fraud Detection ● It is important that credit card companies are able to recognize fraudulent credit card transactions so that customers are not charged for items that they did not purchase. ● Similar application is used in credit providers end as well.
  3. 3. Objective
  4. 4. Technology Stack ● Cassandra ● Scikit / TensorFlow / PySpark ● Kafka ● Flask
  5. 5. Kafka ● Horizontally, distributed & scalable streaming platform. ● Highly Available ● Commodity hardware
  6. 6. Cassandra ● Horizontally, distributed scalable storage. ● Highly available ● Commodity Hardware
  7. 7. Solution
  8. 8. Understanding Dataset ● This dataset presents transactions that occurred in two days, where we have 492 frauds out of 284,807 transactions. The dataset is highly unbalanced, the positive class (frauds) account for 0.172% of all transactions. ● Due to confidentiality issues, we cannot provide the original features and more background information about the data. Features V1, V2, ... V28 are the principal components obtained with PCA, the only features which have not been transformed with PCA are 'Time' and 'Amount'. ● Feature 'Time' contains the seconds elapsed between each transaction and the first transaction in the dataset. The feature 'Amount' is the transaction Amount, this feature can be used for example-dependant cost-sensitive learning. Feature 'Class' is the response variable and it takes value 1 in case of fraud and 0 otherwise.
  9. 9. Study Materials ● Complete Data Science Code - https://github.com/zekelabs/data-science-complete-tutorial ● Credit Card Project Code - https://github.com/zekelabs/ai-project-fraud-detection ● Credit Card Project Videos - https://www.edyoda.com/course/1432 ● More AI courses - https://www.edyoda.com/category/artificial-intelligence
  10. 10. Edyoda Journey
  11. 11. Contact Awantik Das awantik@edyoda.com www.edyoda.com