Deep Learning Based Approach for Milk Quality Prediction
Saumya Kumari, Mahendra Kumar Gourisaria, Himansu Das, Debajyoty Banik
Abstract
This study aimed to find out what factors have the biggest impact on milk quality. We looked at 9 different variables including pH, temperature, taste, odor, fat, turbidity, color, and grade to see which ones were the most important. We used the method called Principal Component Analysis (PCA) and found that temperature and color were the top factors affecting milk quality, with over 95% contribution to PCA-1 and PCA-2. We also divided the milk samples into three grades - low, medium, and high - and used a machine learning algorithm called Artificial Neural Network (ANN) to classify the milk samples. We found that the ANN model was able to classify the milk samples with high accuracy (0.9988) and was more accurate and stable compared to other methods. Other advanced deep learning algorithms could have been used to improve the results further.