Achieved 98% accuracy on CEDAR dataset by developing a handwriting recognition tool using classification techniques such as logistic regression, decision tree, neural networks, and support vector machine to compare handwriting of multiple writers ◦ Attained accuracy of 86% on CEDAR dataset by designing generative models and probabilistic graphical models to learn explainable features for handwritten image pair labeled from same or different writer
fg1804/Forensic-Handwriting-Recognition
Handwriting recognition tool using classification techniques such as logistic regression, decision tree, neural networks, generative models, probabilistic graphical models, and support vector machine
Jupyter Notebook