Machine Learning for Dementia and Alzheimer’s Disease Prediction

Dementia is a progressive neurological disorder primarily affecting cognitive functions, leading to significant impairment in memory, thinking, behavior, and daily functioning. Timely identification of dementia is crucial to facilitate early intervention and improve patient outcomes. Most of the traditional clinical diagnostic approaches are unable to determine the dementia in the early stage. In this study, we have proposed a machine learning-based dementia prediction method to predict dementia at an early stage. Here, to enhance the prediction performance of the machine learning models, we have proposed a relation-based imputation technique to successfully impute the missing values of the dataset. In this work, six different machine learning models are deployed to find out the best machine-learning model for the proposed work. The experimental results exhibit that the prediction accuracy of all the prediction models increased significantly when the proposed imputation technique is applied to impute the missing values of the input dataset. The experimental results also established that Random Forest produces the best prediction accuracy compared to the other machine learning models.