Pharmaceutical research and development requires substantial investment, with costs ranging from $640 million to over $3 billion per compound and development timelines extending beyond five years. A critical safety concern in drug development is cardiotoxicity, specifically the inhibition of the hERG (human Ether-à-go-go-Related Gene) potassium channel, which can lead to potentially fatal long QT syndrome. While molecular dynamics simulations can predict drug-hERG interactions, they are computationally expensive and inefficient. This thesis investigates machine learning as a high-throughput alternative for predicting hERG inhibition. Eight supervised learning classifiers were developed and evaluated: Bernoulli naive Bayes, k-nearest neighbors, logistic regression, random forest, histogram-based gradient boosting, support vector classifier, voting ensemble, and multilayer perceptron. The models were trained on approximately 10,000 molecules from the Therapeutic Data Commons. The support vector classifier achieved the best overall performance with 81% accuracy, precision, and recall, and a Matthews correlation coefficient of 0.63. The multilayer perceptron and random forest classifiers also demonstrated strong performance, with all three models achieving comparable results within a few percentage points. All models significantly outperformed random chance, with the worst-performing models still achieving 70% accuracy. These results demonstrate that supervised machine learning provides a feasible and computationally efficient approach for predicting hERG inhibition, potentially accelerating drug development while reducing costs and safety risks.