Spurious currency detection refers to the identification of fake currency. It involves unauthorized production of counterfeit notes that look much like real currency. Therefore, using technologies like machine learning we can achieve high accuracy for the detection of fake currency. Counterfeit currency poses a serious risk to financial stability and security. We researched upon technologies such as KNN, decision tree, gradient boosting by implementing machine learning and computer vision on various attributes of currency and performing wavelet transformation on the curated dataset. Parameters such as kurtosis, skewness, variance are taken to evaluate for any outliers or redundancy in data. This paper highlights the role of machine learning algorithms in improving detection accuracy and reducing false alarms. We evaluated algorithms mostly based on machine learning (KNN, logistic regression, pruning tree and XG support). Moreover, the performance was evaluated based on some multivariate metrics such as accuracy, f-score, specificity, sensitivity, precision. By implementing machine learning algorithms on the dataset, we found that KNN provides the highest false positive performance accuracy among different models.