Alzheimer's Disease (AD), as the most devastating neurodegenerative disease worldwide, has reached nearly 10 million new cases annually. Current technology provides unprecedented opportunities to study the progression and etiology of this disease with the advanced in imaging techniques. With the recent emergence of a society driven by big data and machine learning (ML), researchers have exerted considerable effort to summarize recent advances in ML-based AD diagnosis. Here, we outline some of the most prevalent and recent ML models for assessing the progression of AD and provide insights on the challenges, opportunities, and future directions that could be advantageous to future research in AD using ML.