Parkinson's disease (PD), a neurodegenerative disease with symptoms hard to distinguish from other disorders, affects millions of people globally. Among the symptoms of PD, changes in gait have been used as a primary diagnosis factor. Symptoms of PD include bradykinesia, tremors, depression, hallucinations, cognitive decline, and falls. This study presents a dataset that records data on PD patients who experience freezing of gait, including data for medication in the “on” and “off” states. Classification is applied to two problems relating to this PD data: first, to distinguish PD patients from healthy individuals, and second, to determine effectiveness of medication for freezing of gait. Among the classifiers considered, Multilayer Perceptron, K-Nearest Neighbors, Random Forest, and Support Vector Machine obtain the best results when applied to these problems.