Applying logistic regression and Deep neural network for the Ekush dataset and building a binary classification model that can predict male/female from Bengali handwritten digits with different hyperparameter settings. The dataset used is Ekush dataset, a dataset which contains pixel value of about 30,830 images. It is a labelled dataset with informative tags which are manually annotated. It contains information about male and female handwritten digits which will help to do detect the male and female writing. This is a binary classification problem.
Dataset Link: https://shahariarrabby.github.io/ekush/#home