Several issues are there to prevent the traditional classifiers from getting an acceptable performance level while learning from multi-class problems. One of the core issues is the imbalanced distribution of samples, which in unification with incongruous optimization benchmarks and data overlapping phenomena dramatically decrease the performance of the underlying classifier. The joint impact of imbalance distribution and sample overlapping near the class boundaries compromise over the classifier performance beyond the expectation level and become even more challenging with the increasing number of classes in the multi-class environment. Learning from imbalanced data studied extensively in the research community, however, the overlapping issues and the co-occurrence impact of overlapping with data imbalance have received comparatively less attention, even though their joint impact is more thoughtful on classifiers\' performance. This paper introduces SVM++, a modified version of Support Vector Machine (SVM) to enhance the learning from the complex scenarios of multi-class problems with the imbalance and overlapping data with a shared attribute in the overlapped region. The proposed techniques is implemented using three steps. In the first step, an algorithm is designed to divide the training set into the overlapped and non-overlapped region at the data preprocessing level. In the second step, the overlapped data further filtered into the Critical-1 and Critical-2 region. In the third step, the SVM++ transforms the Critical-1 region sample, sharing similar characteristics into higher dimensional space by altering the SVM kernel mapping function based on the mean of the maximum and minimum distance. For the experiment, we use 30 real datasets with varying imbalance ratio and the overlapping degree to compare the novelty of the SVM++ with the existing classifiers. Experimental results highlight the superiority of the proposed SVM++ on a collection of benchmark datasets to its standard counterpart classifiers.