The objective of this study was to prove the general applicability of an GC-TD-MS machine for analyzing different types of coffee beans like roasted coffee aroma considering the dependency of different roasting levels usually we analyzed the gas chromatograms obtained for the four different roast levels (from A to D) of coffee beans, Robusta coffee beans from Guatemala considering the raw coffee and yeast treated coffee and finally we concluded our study for coffee beans obtained from four different geographic origin coffee beans from Africa. This thesis proposes methods to analyze volatile organic compounds obtained from GC-TD-MS machine. First, a general data analysis methodology was proposed to analyze the data set obtained from the TD-GC-MS. The proposed methodology constitutes of five steps Data collection, Data cleaning, explanatory data analysis, Building supervised classification models and in the final phase we concluded the overall classification accuracy obtained. Raw Data obtained from TD-GC-MS, data cleaning is performed by using MS-access, MS-Excel and important compounds chosen for further analysis, in the explanatory model we used bar chart to plot mean values to measure the tendency of roasted coffee beans and PCA is used in explanatory phase to find pattern in discriminating roasted coffee, coffee stored under different temperature, finally used in discriminating different geographic coffee. In classification modeling phase supervised methods such as KNN and LDA are used to discriminate different coffee beans stored under room temperature and refrigerator temperature in the final phase we concluded i.e. specific process work of TD-GC-MS tool system for measuring gaseous compounds of different specimens. And it's practical and beneficial of the research for the coffee preparation.