Machine Learning in FinTech

This project aims to supplement Allied Payment Network's PicturePay OCR with image classification using machine learning. Customers can submit bill photos via the mobile app, and any OCR failures are handled by the support team.

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

• Machine learning
• FinTech
• bill classification
• neural networks

leaf Follow

Uploaded on Dec 22, 2023 | 11 Views

Presentation Transcript

1. Machine Machine Learning in Learning in FinTech FinTech Project Team #09 Adam DeWitt Andrew Hendrickson Ashwin Rajnish Bricks Hudson Jacob Gottschalk

2. Client: Allied Payment Network Local Fort Wayne Payment technology for banks and credit unions Bill Classification with Neural Networks Goal: Supplement APN PicturePay OCR with image classification Customers submit bill photos via mobile app OCR failure is handled by support team

3. System Model

4. Sequence Model

5. Convolutional Neural Network Overview

6. Feature Learning: Kernel Convolution

7. Feature Learning: Max Pooling

8. Fully Connected Layers Spatial Y Spatial X Flatten

9. Roadblocks and Risks Restricted dataset access Limited workstation access Model training time Model accuracy improvement could be difficult Overfitting

10. DevOps

11. About Preprocessing Improved feature learning and computation time Edge and corner detection Transformation

12. Functional and Non-Functional Requirements Functional Requirements Nonfunctional Requirements Image input Classification in less than 5 seconds Classify image (recognize top 100 billers) 24/7 online accessibility Output classification with confidence score Classification accuracy at 70% or better Product must fit into their current modules

13. Technologies GitHub NodeJS PyTorch Machine Learning Library for Python Boto3 Amazon S3 Devt. Kit

14. Project Schedule October November December January February March April Complete DevOps pipeline Establish NN model with at least 50% accuracy +5% model accuracy +5% model accuracy +5% model accuracy +5% model accuracy MVP deliverable

15. Glossary Artificial Neural Network - layered machine learning system inspired by an account of neuron activation in the brain. Confidence - output indicating a machine learning program s amount of certainty in its classification; may be expressed numerically or verbally. FinTech - financial technology , the use of new technology to compete with traditional finance services. GitHub - internet hosting service for software development and version control. Machine Learning - field of understanding and building methods which leverage data to improve performance on some set of tasks. Minimum Viable Product - product with minimum features to be built upon before release. Model - file which has been trained to recognize patterns.