Research Projects
Search & Explore
November 2021 - late 2022
An Adaptive and Dynamical Neural Network for Machine Remaining Useful life Prediction
Supervisor: Dr. Xiaoli Li, Dr. Zhenghua Chen and Dr. Ruibing Jin
Accepted by: IEEE Transactions on Industrial Informatics [ISSN: 1551-3203]
May 2021- August 2021
Theory of Artificial Neural Networks
Supervisor: Professor Roman. Kuc
Research Content: The purpose of this research is connecting YOLO & CLIP to create your own data set through your own videos for retraining, research and expansion. And this project is an attempt to improve the quality of the data, using the elements in the own video to make your own data set.
Video zero-shot generation model, based on YOLO & CLIP by inputting video of any size and any common format (of course, a picture set is also possible), after YOLO cropping and attaching an initial tag, CLIP expands the picture in an optional database of various images of this video content can be obtained.
Use innovative methods to create datasets from videos. Load the video, use YOLO to process each frame of the video, and save the frame image. Collect the pictures recognized by YOLO and save in the corresponding label folder. Use clip to search for similar picture features and IDs in the chosen database. Download similar pictures in the database to the label directory to expand the styled dataset of the own video.
August 2020 - November 2020
Futures and Stock Prices Properties and Predictabilities
Supervisor: Professor Alexei. Chekhlov
Research Content: In financial practice, most homogeneous strategies have similar decision-making results, but the sensitivity of the individual is different. This research explores and analyzes the performances of eight future markets from China and the United States. It adopted the Trend-Following strategy, and analyzed the result with the aid of charts and excels.
As the group leader, led the group to finish the research report and gained the score of A; completed the summary of paper and the paper will be published soon
Be responsible for algorithm; used the Kelly formula to calculate the profit-loss ratio and winning rate in real time to adjust the position
Be responsible for the programming on Matlab; used two variables to go through a traversing method and found the two numerical values which should be used in the following process.
Made comparisons between the in-sampleand out-of-sample results of each future market
Research as: Minor Degree Thesis with A.
Repo: https://github.com/Lemutisme/Prediction-of-Future-Market
June 2020 - April 2020
Research on Internet Finance Innovation and Traditional Bank Transformation Based on Big Data
Research Content: In recent years, China's big data and information technology have been widely used. Under this situation, the traditional financial industry has been given rapid development opportunities, and the level of traditional banking business has gained a greater degree in the context of the development of Internet finance. improve. Traditional domestic banks are gradually transforming to Internet financial banks, but for traditional banks, this transformation and upgrade is not only an important opportunity, but also will face this unprecedented test.
Published on WOP in Education, Social Sciences and Psychology [ISSN 2515-2904]
March 2019 - July 2020
Research Experiment on Color Psychological Mechanism
Supervisor: Professor Xinguang Lyu
Research Content: It assumes that color is a kind of information and aims to establish a new analytical theoretical framework-consciousness symbolism, to study the influence of color on human behavior and to reach preliminary conclusions in line with actual observations, solve the simple principle of color psychological influence mechanism and apply it to advertising , packaging, marketing, space design and other fields.
Independently completed investigation, research, data analysis, experiment, and thesis writing.
Conducted dialectical analysis and discussion, implemented VGG16 to carry out the color consciousness simulation growth experiment, and used the machine learning method to initially establish the research framework of consciousness symbolism.
Complete the thesis "The Psychological Influence Mechanism of Color Consciousness Symbols", and carry out a complete extension on this basis.
Research as Major Degree Thesis with A.
December 2016 - February 2017
Difficulties and Decisions in the Development of Rural E-commerce in China
Research Content: In recent years, China’s rural e-commerce is developing rapidly. E-commerce has played an important role in promoting the transformation and upgrading of the agricultural industry, promoting local economic development, and implementing targeted poverty alleviation. However, its development is facing many problems. It also requires policy inclination and market guidance to seek solutions on the chain, logistics chain, and talent chain.
Published on Consumer Guide [ISSN 1672-5719]
If you have questions about my experience, don’t hesitate to reach out.