LumpySkinDisease_DataHub

Welcome to the Lumpy Skin Disease Data Hub Lumpy Skin Disease (LSD), or nodular skin disease, is a viral disease that mainly affects cattle, particularly cows. It is caused by a poxvirus of the family Poxviridae. The disease gets its name from the nodules or lumps that form on the skin of infected animals. Typical symptoms of LSD include the appearance of nodules on the skin, fever, eye discharge, swelling, and pain in the affected areas. In some cases, the disease can lead to reduced milk production, weight loss and impairment of the animal's general condition. The disease is spread primarily through direct contact between infected animals or through vectors such as mosquitoes and flies. It can have a significant impact on the livestock industry, causing economic losses due to reduced milk production, animal mortality and trade restrictions imposed on affected countries. Prevention of LSD relies mainly on isolation of infected animals, vaccination and control of vectors that transmit the disease. Here you will find how to train a deep learning/machine learning model for LSD-related image classification. This repository is structured using the following folders: code - lumpy_multi_DL_bal.ipynb: This notebook contains code and documentation of a deep learning implementation for image classification for LSD disease. The wording "bal" suggests that the dataset used for training the model is balanced, that is, it contains a similar number of samples for each class. - lumpy_multi_DL_imbal.ipynb: This notebook is similar to the previous one, but uses an unbalanced dataset for model training. This means that some classes may be overrepresented relative to others in the dataset. - lumpy_multi_ML_bal: This notebook contains code and documentation for an implementation of machine learning algorithms for classification of LSD disease images. The wording "bal" suggests that the dataset used is balanced. - lumpy_multi_ML_imbal: Similar to the previous point, but uses an unbalanced dataset for model training. data-processed - lumpy_balanced.zip: This folder contains a set of images related to LSD. The wording "balanced" suggests that the dataset has been balanced, i.e., it contains a similar number of samples for each class of interest, which are necessary for training and evaluation of machine learning or deep learning models. - lumpy_unbalanced.zip: This folder contains a dataset similar to the first, but with the difference that the dataset is unbalanced. This means that some classes of interest may be overrepresented relative to others in the dataset. Within the archive, one may find the same types of data files present in the case of the balanced dataset.