ML Visualizer

Through an interactive graphical user interface that facilitates understandable, transparent, and effective model training for practical applications like counterfeit banknote detection, this project significantly contributes to closing the gap between sophisticated machine learning algorithms and user accessibility. The study presents a Machine Learning Graphical User Interface (ML GUI) frame- work developed for both counterfeit currency recognition and general-purpose classification tasks using statistical parameters including variance, skewness, kurtosis, and entropy. Without the need for complex coding skills, the system offers an intuitive environment for data preparation, feature visualization, model training, and performance evaluation. It incorporates multiple algorithms—Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), K-Nearest Neighbors (KNN), Linear Regression (LR), Logistic Regression (LogR), and Multilayer Perceptron (MLP) to ensure diverse and reliable classification. Model effectiveness is evaluated using accuracy, precision, recall, and F1-score, while visual tools such as confusion matrices enhance interpretability. Experimental analysis shows that Decision Trees and Random Forests deliver high accuracy and transparency, SVM performs well on linearly separable data, and MLP effectively models complex non-linear relationships. Developed using Scikit-learn and TensorFlow, the GUI framework is scalable, educationally valuable, and adaptable for real-time fraud detection, making machine learning more accessible for both learners and professionals.