Global warming, caused by Carbon Dioxide (CO2) emissions from hard-to-abate industries, necessitates CO2 capture as one of the promising alternatives. Post-combustion CO2 capture (PCC) can be retrofitted for industries where chemical absorption using an amine-based solvent is an efficient method for CO2 capture. An essential factor for CO2 absorption in an amine mixture is the study of CO2 solubility and CO2 + Amine physicochemical properties to identify the more efficient solvents. Often experimental data are available in the literature, but precisemodeling is necessary to use such data for further utilization in a process-plant model. In this work, a blended solvent aqueous 2-amine-2methyl-1-Propanal (AMP) with Piperazine (PZ) and 1-(2-aminoethyl) piperazine (AEP) have been identified as a potential candidate for PCC application. The fundamental properties such as the density, viscosity and CO2 solubility in AMP + PZ and AMP + AEP blended solutions have been predicted by employing machine learning to get an exceptionally efficient model. This data-mining technique efficiently calculates solubility, density, and viscosity by using 691, 559, and 559 experimental data sets respectively, from the literature of AMP + PZ. Similarly, 120 data sets of solubility have been collected from the literature for AMP + AEP. Random Forest regression, Artificial Neural Network (ANN), and XGBoost regression algorithms are adapted to predict CO2 solubility in aqueous AMP blended solution. The effectiveness of these models has been confirmed through statistical analysis, leading to the conclusion that XGBoost outperforms the other two approaches. Moreover, an identical approach has been employed to develop a predictive model for the density and viscosity of AMP + PZ solvent. The outcomes imply that the machine learning algorithm performs better with experimental data and may be utilised as an effective simulation tool.