Machine learning SDK for beginners

Abstract

This is a development project that aims to develop a tool that helps to build machine learning related features easily and quickly. A beginner without any background of machine learning or less programming experience are the main target users. The core functions are building a classification model and its integration with the mobile app and server. In this project, I used Python to build a tool to train a machine learning model and to do server integration. Also, I developed an iOS example app that integrating with the trained model and server written in Swift. It consists of multiple features such as real-time object. There are many external libraries available online that help to train a model. I studied more than 20 open source libraries and service providers. I integrated libraries which are suitable for beginners and easy to use. An application that is available on the App Store used this project to develop the classification model and build the server. Also, I won the most innovative team in PwC’s Data-Lympics 2019 with the help of this project. As a result, this project is successful and can be used by the public after the documentation of the tool is completed.