UROP Proceedings 2020-21

School of Engineering Department of Computer Science and Engineering 110 Practical ML-based Mobile Applications Supervisor: CHATZOPOULOS Dimitrios / CSE Student: FERDY / COMP Course: UROP1100, Fall Artificial Intelligence has been one of the biggest topics in society nowadays. The benefit of using Artificial Intelligence, such as machine learning, in everyday life is substantial and therefore all industries are in an arms race to incorporate AI to their products. The fast growth of AI startups around the world showing how companies try to maximize user experience by analyzing the input data entered by the user, record their usage pattern and user feedback. The proportion of people who use mobile phones also increases every year. Thus, the practicality of mobile development combined with sophisticated machine learning algorithms can offer users better experience and a wide range of innovative products. Practical ML-based Mobile Applications Supervisor: CHATZOPOULOS Dimitrios / CSE Student: TOH Magdalene Youjun / COMP Course: UROP1100, Fall Meal planning is important to reduce food wastage and ensure one has a balanced diet. However, one may run out of recipe ideas if one cooks frequently. While there is an abundance of recipes on the internet, it can be time-consuming to filter through them to discover recipes that are suitable. Hence, the purpose of this project is to produce a mobile application that gives personalized recipe recommendations to its users. The recommendation system utilizes content-based filtering to recommend recipes based on user preferences such as preferred preparation time, equipment, and types of ingredients. After candidate generation, the recipes will be further ranked according to the similarity of the ingredients in the recipes to the food supplies that the user already has. Practical ML-based Mobile Applications Supervisor: CHATZOPOULOS Dimitrios / CSE Student: WONG Yeuk Nam / IS Course: UROP1100, Fall Machine learning has been widely used in daily lives for more accurate predictions and higher working efficiency. With the developing technology on optical character recognition and the high bookkeeping cost among companies, this project aims at developing a mobile application that detects total price from receipts for automatic bookkeeping records. This project has selected Android Studio as its application development environment. To develop the application, different online resources and cloud services have been evaluated. Given the limitations with time and app development knowledge, the approach is to build additional features on existing app resources that align with the basic features including the use of smartphone camera. After recognizing numbers from receipts, the application should be able to insert records into a data table that facilitates further processing such as analysis of monthly expenditure. The application is still under development and not ready to work smoothly on mobile yet because of the incapability between the original application and the additional features.