Different firms use completely different analysis tools for foretelling and also the main aim is that the accuracy with that they predict that set of stocks would yield the utmost quantity of profit. This paper offers a quick introduction to varied techniques used for prediction so it's simple for buyer/seller to make a decision. | ISSN:2249-5789 Roshani Gandhe et al, International Journal of Computer Science & Communication Networks,Vol 6(1),13-17 Performance Analysis of Stock Market Prediction Techniques Roshani Gandhe Student, Computer Science & Engg. Priyadarshini Bhagwati college of Engg, Nagpur India Manoj S. Chaudhari Prof & HOD, Department of CSE Priyadarshini Bhagwati college of Engg, Nagpur India Abstract Predicting exchange accurately has invariably intrigued the market analysts. During the past few decades varied machine learning techniques are applied to check the extremely random nature of exchange by capturing and victimization repetitive patterns. Different firms use completely different analysis tools for foretelling and also the main aim is that the accuracy with that they predict that set of stocks would yield the utmost quantity of profit. This paper offers a quick introduction to varied techniques used for prediction so it's simple for buyer/seller to make a decision. Keywords — Neural network, Regression, Time series model I. Introduction Time series involves learning the observations chronologically at regular intervals. Global crashes ar headed by native crashes and don't happen all of a sudden .The increase within the use of temporal information has initiated varied development and analysis efforts within the space of knowledge mining. Statistic is a very important category of temporal information observations and may be simply obtained from scientific and monetary applications (. Weekly sales totals, daily climate predictor, costs of mutual funds and stocks).Time series is a very important highdimensional information sort. In contrast to static information, the statistic of a feature comprise values modified with time. statistic information are of interest attributable to its generality in varied areas starting from science, engineering, meteorology, business, stock market, economic, health care, to government. Time series is predicts future .