Easy to use class balanced cross entropy and focal loss implementation for Pytorch

Implementing contractive auto encoder for encoding cloud images and using that encoding for multi label image classification

Different Loss Function Implementations in PyTorch and Keras

Implementation of gradient descent from scratch with binary cross entrophy loss

This project involves building an Artificial Neural Network (ANN) for predicting customer churn. The dataset used contains various customer attributes, and the ANN is trained to predict whether a customer is likely to leave the bank.

What Happens if We Use a Mean Squared Error Loss for Binary Classification?

Predict which Tweets are about real disasters and which ones are not Microsoft DeBERTa

Implementation of some basic Image Annotation methods (using various loss functions & threshold optimization) on Corel-5k dataset with PyTorch library

Job Prediction given job description and skills

NLP Model for Spam/Ham Mail Classification

Simple neural network to detect offensive language usage

Utilize autoencoders for anomaly detection and customer credit risk evaluation

A collection of tensorflow projects.

Utilized CNN models to classify images of mountains and forests, treating mountains as the positive class and forests as the negative class. We compare the performance of a pre-trained model, a custom CNN model, and a CNN model with data augmentation.

Alternative loss function of binary cross entropy and focal loss

Used a Multilayer Perceptron (MLP) neural network to detect COVID-19 in lung scans.

A Generative Adversarial Network for the generation of new synthetic art.

Mining gold from implicit models to improve likelihood-free inference, example for ROLR and RASCAL.