One of the most prevalent neurological illnesses, Parkinson's disease (PD) mainly affects the motor system of the brain's core system. In actuality, PD is marked by speech impairment, tremors, muscular rigidity, and gait inaccuracies. Even though the primary symptoms of Parkinson's disease cannot be clearly distinguished from those of other disorders, a definitive diagnosis of the condition is often dependent on several neurological, psychiatric, and physical studies. As a result, several machine learning-based automatic diagnostic assistance systems have lately been used to aid in the assessment of PD patients. One of the most difficult medical issues at hand today is the automatic diagnosis of early Parkinson's disease using feature data sets. Such datasets contain several characteristics that are either worthless or plagued by issues like noise that hinder learning and add to required computing load. This article suggests a hybrid feature selection algorithm built on an enhanced correlation method with Bootstrap to increase the effectiveness of feature selection to ensure most accurate performance of the classier. By combining the finer elements of filters and wrappers, such an algorithm finds the ideal subset of features by removing a majority of noisy or unrelated information. Select optimal features from overall data features is also specific issue behind implementation of feature extraction In order to overcome such problems, we have proposed an adaptive random forest classifier method that uses ensemble feature selection technique for better information gain (IG), improved correlation (IC) and gain ratio (GR). Also,s it seeks to solve the class imbalance problem by applying bootstrap re-sampling for medical data. According to the evaluations in terms of accuracy, precision, recall, and F-Score, the developed model has been found to be more efficient than conventional methods. The analysis of experimental results indicates better accuracy of the proposed framework (88.3% accuracy) as compared to other techniques.