Sustainable Development Goals
Abstract/Objectives
As people pay more attention to their privacy, in order to build a secure and robust cloud infrastructures for processing data to make services better, in 2016 Google introduced a new approach called Federated Learning. Federated Learning enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on device. Federated Learning allows for smarter models, lower latency, and less power consumption, all while ensuring privacy. Also, in addition to providing an update to the shared model, the improved model on device can also be used immediately, powering experiences personalized by the way users use their phones. The goal of this internship is to do research on Federated Learning and to build a simple model.
Results/Contributions

There are two harvests in this internship, one is self enhancement, and another is cognition of future career development. For self enhancement, I gained a general concept of Federated learning, learned how to use Ubuntu system, and tried using FedAvg algorithm to build a simple model. Also, I realized that self-learning skill is so important not only when being a student but also in working. Especially in the era of big data, capabilities must keep pace with the times. There are many ways to gain new knowledge, for example reading paper, online materials, and online courses. No matter what kind of ways you learn, finding a way which suitable for yourself and cultivate a ‘Lifelong Learning’ attitude are the most important. Through this internship, I can find the difference between academia and industry. Institute for Information Industry is close to academia, and they focus on research instead of application. With this experience, I realized that due to my personal interest if having other chances to be an intern or seeking for a job, I will apply for another type of industry.   

Keywords
Federated LearningMachine learningPytorch