This research addresses a prominent concern in contemporary society, non-alcoholic fatty liver disease (NAFLD) arises from an accumulation of fat in the body, posing a substantial health challenge. In essence, obesity-related liver illness is more harmful since it shortens people's lives. The liver releases fats, such as triglycerides and hyperlipidemia. These are a few of the lipids that the liver has evolved for the bloodstream. The liver may be impacted by the slowing down of blood flow that occurs when these fats are consumed in excess. Inflammation and damage to the liver are possible outcomes if fat storage is found in the liver cells. Therefore, identifying liver diseases is essential to minimizing their detrimental effects. To create a good prediction model, a few machine learning methods are being studied the proposed CROSS VALIDATION and boosting techniques system is intended to treat non-alcoholic fatty liver disease. We thoroughly suggested the Randomized search CV and Grid search CV algorithms in addition to other widely used models. According to the experimental results, the suggested architecture generally improves the accuracy of the disease predictions, ensuring a high level of model resilience and robustness