The TMVA (Toolkit for Multivariate Data Analysis) is a ROOT-integrated environment designed for multivariate classification and regression analysis, particularly useful in high-energy physics. Version 4 of TMVA introduces multivariate regression capabilities and enhanced data handling flexibility. The toolkit includes a variety of multivariate classification algorithms, such as rectangular cut optimization, projective likelihood estimator, multidimensional likelihood estimator, and boosted decision trees. TMVA provides user-friendly interfaces for training, testing, and performance evaluation of these methods, along with visualization tools. The toolkit supports data preprocessing, probability density function estimation, optimization, and fitting techniques. TMVA is open-source and available under the BSD license, with contributions from various institutions and researchers. The documentation covers installation, quick start, detailed usage, and examples for both classification and regression tasks.The TMVA (Toolkit for Multivariate Data Analysis) is a ROOT-integrated environment designed for multivariate classification and regression analysis, particularly useful in high-energy physics. Version 4 of TMVA introduces multivariate regression capabilities and enhanced data handling flexibility. The toolkit includes a variety of multivariate classification algorithms, such as rectangular cut optimization, projective likelihood estimator, multidimensional likelihood estimator, and boosted decision trees. TMVA provides user-friendly interfaces for training, testing, and performance evaluation of these methods, along with visualization tools. The toolkit supports data preprocessing, probability density function estimation, optimization, and fitting techniques. TMVA is open-source and available under the BSD license, with contributions from various institutions and researchers. The documentation covers installation, quick start, detailed usage, and examples for both classification and regression tasks.