WriteVision (Digit Recognizer)

Google Docs - WriteVision Report

Models

• FCNN Model

Project Overview

The FCNN classification models were developed using various optimization algorithms, including SGD, VGD, SGD(momentum), SGD(NAG), AdaGrad, RMSProp, and Adam. These models were designed to classify digits and achieved a best performance accuracy of 98.42%. Additionally, a detailed report was also created.

Tools Used

• Tensorflow
• Keras
• Scikit-learn
• Pandas
• Matplotlib
• Python