assigning premium to the insurance contract in iran mostly has based on some old rules have been authorized by government, in such a situation predicting premium by analyzing database and it’s characteristics will be definitely such a big mistake. therefore the most beneficial information one can gathered from these data is the amount of loss happens during one contract to predicting insurance loss of the new arrival record into system. in fact, by running machine learning algorithm three goals are followed: • defining risky cases • extracting out the non effective attributes of the cases • predicting insurance loss by considering the results of these outputs, before assigning insurance contracts, one insurance company can achieve the lowest level of losses during a year.