This study emphasizes on high-risk investment in Indian Stock Market especially on Small Cap Stocks. It will help the high-risk taking investors to minimize their risk and can get a positive return. In this study seventy-seven small cap stocks have been taken into consideration based on Automobile, Bank, Biotech and Pharmacy, Electronics, Finance, Food and Beverage, Information Technology and Metal & Chemical sectors. From each sector eight stocks have been taken and been used to develop a tailored portfolio. Altogether ten portfolios have been developed taking one stocks from each sector. The first attempt made here was to develop ten different portfolios with the help of machine learning through random number generations with the help of python. Later, concept of minimum standard deviations of each stock was taken and four portfolios are made based on “lower the standard deviation, lower is the risk” and a comparative analysis was done with the machine-driven portfolio. Comparative analysis was performed, and it was found that though machine driven portfolios are giving good results, but man-made portfolios are better than machine driven portfolios. Random Forest regression, Convolution Neural Network and LSTM(RNN) has been taken for forecasting of both type of portfolios. Errors were measured with MADP, MSEP, RMSEP and MAPEP. The historical prices of stocks are taken from Yahoo finance, NSE and BSE.