Milk is an important naturally occurring beverage high in nutrients and is essentially a part of our daily lives. Water is frequently added to milk to increase its volume and raise the seller’s profit. Although there are various techniques for determining the dilution level in milk, they have certain drawbacks in terms of cost, sensitivity, accuracy, and how quickly findings may be obtained, among other factors. The robust Random Forest Regression approach was used in this study to forecast the milk dilution series. This paper shows how optics and machine learning can be combined in novel ways to automate and improve dilution analyses in scientific settings. The adaptability of the Random Forest algorithm to different concentrations and environmental circumstances suggests that it could find wider uses in bioinformatics and analytical chemistry.