Due to the developing quantity of digital payments, the monetary strain of credit-card fraud is turning into a great assignment for fi nancial establishments and service providers, for that reason forcing them to always enhance their fraud detection systems.E-commerce is becoming popular with online shopping, banking , financial institu tions and government. Fraudulent activity is leaving in many areas of business and our daily lives. These activities are most common in telecommunications, credit card fraud detection, network intrusion, fi nance and insurance, and scientific applications.Billions of dollars are lost every day due to the increase in the number of credit card trans actions used online.The customary fraud detection model depends on a static guidelines based framework, additionally alluded to as a cre ation or master framework. Despite the fact that these frameworks have been compelling for quite a while, a portion of their significant disservices makes them unsatisfactory for present-day advanced situ ations. Static guidelines put together framework is vigorously needy with respect to human work. In any case, normally, top experts are costly. Also, their work requires some serious energy. Likewise, even top specialists make rules dependent on their insight, abilities, and experience, which are constantly constrained. Such principles can develop to gigantic sizes and get so mind-boggling that it’s about outlandish for an outcast to comprehend them when required. Like- wise, making another standard and actualizing it takes some time when done by hand. Designing powerful and efficient fraud detec tion algorithms is the key to reducing transaction losses. Numerous approaches to the detection of fraud have been implemented. Our Proposed solution is Based on the logic regression algorithm using 3 method:The vanilla logistic regression,The logistic regression with oversampling and the logistic regression with balanced weights and we will see which method will give us the best false negative and false positive rate .