Background of the Problem and Motivations:
Privacy is a very important in terms of most of the population today. A company or organization that can provide privacy to their users is something that is not seen as a benefit, but something that is to be expected. That being said machine learning (ML) and artificial intelligence (AI) is also a hugely popular technology for these same companies. ML and AI can help companies customize their products to help solve customer problems and identify customer wants and needs. This can be an issue however simply due to the fact that generally machine learning is trained on large amounts of sample data that is provided by the user. This creates a conflict between keeping user data private and having enough data to provide useful results that can then be used to implement changes in the technology.
Previously, centralized learning was implemented as a way of gathering user data. All of a user’s data would be uploaded to a single server and the data would then be used for training models or other purposes. However, more recently users are more aware of their data and less willing to share their information with others. Additionally, data protection and privacy laws have been or will be implemented that protect a user’s data. These laws and regulations will disallow direct access to certain types of data, limit the amount of data that can be used, or the time that the data can be kept.
Federated learning also known as collaborative learning, has been shown to rectify the issues that are involved with centralized learning. Federated learning allows for information to be taken from the user in a way that the user’s privacy is not violated. The user’s device will obtain a statistical model from a centralized server. The model will then be trained using the information held within the user’s device. Finally, the altered model will be returned to the centralized server. All of this occurs without the user’s personal data leaving their device. Gathering information in this or similar manners will be crucial as organizations must adhere to privacy laws.
In this paper I will look into the downfalls of using centralized learning. I will also explore machine learning techniques, namely federated learning, that will best preserve the privacy of the user as well as providing the researcher, company, organization, or otherwise with useful information. Samples shown in the following will demonstrate how user information can be obtained that is not detrimental to the privacy of the user.
A professional writer will make a clear, mistake-free paper for you!Get help with your assigment
Please check your inbox
I'm Chatbot Amy :)
I can help you save hours on your homework. Let's start by finding a writer.Find Writer