In today’s fast moving world as the demand of the internet is increasing with this cyber-attacks are also increasing and Phishing is one the most common cyberattack among them. Taking a user’s personal information such as credit card numbers, login details, confidential info, and so on through illegitimate activities of using the internet without the user’s knowledge and using it for blackmailing, debiting money from the user’s account, or any other purpose with the wrong intentions is known as phishing. In phishing attack, a phisher or attacker pretends to be masquerade as a known person or organization to the user and sends the mails or messages which contains malicious links in them and these malicious links contains harmful software’s or viruses which steals the user’s computer data, financial data, login credentials such as User ID and passwords, credit card details, etc. Phishing is the most common and dangerous cyberattack which is growing in the today’s world. Nowadays phishers are working smartly they are using the new techniques for creating the malicious links and embeds them in the emails and messages and sends it to the user which looks similar to the trusted mail or message to the user and as soon as the user clicks on the malicious link it redirects the user to the malicious webpage or runs the harmful software in the backend while the user is reading the email or message and takes over the соntrоl оf the user’s соmрuter аnd steаls аll dаtа оf the user’s соmрuter. Due to a speedy development inside the digital commerce generation, the usage of credit playing cards has dramatically extended. In view that credit card is the most popular mode of fee, the number of fraud instances related to it is also rising. As a result, in order to prevent these frauds, we need an excellent fraud detection system that can detect them correctly. We created the idea of credit card frauds in this paper, and we used a variety of device learning methods on an unbalanced dataset, including logistic regression, naivebayes, and random wooden area with ensemble classifiers using the boosting approach. An in-depth analysis of the existing and proposed models for credit card fraud detection has been completed, as well as a comparison of these tactics. So one of a kind classification models are applied to the statistics and the model performance is evaluated on the basis of quantitative measurements which include accuracy, precision, recollect, f1 score, confusion matrix. The realization of out observe explains the first class classifier via schooling and trying out using supervised strategies that offers better answer.