Differential privacy is a formal privacy guarantee that ensures the removal or addition of a single database item does not significantly affect the outcome of any analysis. This guarantees that no individual's privacy is compromised when their data is included in a statistical database. The concept has gained significant attention in the past five years, with many results published in various fields including statistics, databases, theory, and cryptography. This survey provides an overview of differential privacy, including its definition and two basic techniques for achieving it. It also presents applications of these techniques, including algorithms for three specific tasks and three general results on differentially private learning. The survey highlights the importance of differential privacy in protecting individual privacy while allowing for the release of statistical information. Privacy-preserving data analysis, also known as statistical disclosure control, involves releasing statistical information without compromising individual privacy. There are two main settings: non-interactive and interactive. In the non-interactive setting, the curator computes and publishes statistics, while in the interactive setting, the curator interacts with users to protect privacy. Differential privacy is a strong privacy guarantee that is independent of the computational power and auxiliary information available to an adversary. It is not an absolute guarantee of privacy, but rather a statistical property that ensures that the presence or absence of an individual's data does not significantly affect the outcome of any analysis. This makes differential privacy a powerful tool for protecting individual privacy in data analysis.Differential privacy is a formal privacy guarantee that ensures the removal or addition of a single database item does not significantly affect the outcome of any analysis. This guarantees that no individual's privacy is compromised when their data is included in a statistical database. The concept has gained significant attention in the past five years, with many results published in various fields including statistics, databases, theory, and cryptography. This survey provides an overview of differential privacy, including its definition and two basic techniques for achieving it. It also presents applications of these techniques, including algorithms for three specific tasks and three general results on differentially private learning. The survey highlights the importance of differential privacy in protecting individual privacy while allowing for the release of statistical information. Privacy-preserving data analysis, also known as statistical disclosure control, involves releasing statistical information without compromising individual privacy. There are two main settings: non-interactive and interactive. In the non-interactive setting, the curator computes and publishes statistics, while in the interactive setting, the curator interacts with users to protect privacy. Differential privacy is a strong privacy guarantee that is independent of the computational power and auxiliary information available to an adversary. It is not an absolute guarantee of privacy, but rather a statistical property that ensures that the presence or absence of an individual's data does not significantly affect the outcome of any analysis. This makes differential privacy a powerful tool for protecting individual privacy in data analysis.