Abstract
The dexterity of portfolio management and selection of scripts has always been a challenging realm for the researchers. After the advent of the modern computing methods and machine learning environment much of the automation activity has been adopted by the computational finance researchers but still there is room for betterment. The current research is extension of the automated portfolio selection being previously carried out, in which chameleon and dynamic K-Means algorithms were modified through ensembled learning through which a set of scripts are being selected. Current research undertakes the adjustment of the weightage of the selection in the portfolio for the diminution of the risk and amplification / upsurge of the return through the dividend yield and the capital growth by utilizing the binary coded genetic algorithm.