Graduation Thesis 2

Based on the knowledge acquired up to the second year in the Department of Computer Science and Engineering, each student chooses a theme for his or her graduation thesis that interests him/her or that he/she would like to study in depth. Through this, students learn research methods, how to organize results, how to compile a thesis, and how to make a presentation.
Depending on the research theme, part of the research may be conducted abroad.
Students who have not obtained credits for Graduation Research 1 cannot enroll in Graduation Research 2.

　By utilizing the knowledge they have learned so far and working to solve the problems they have set themselves, students will enhance their comprehensive abilities that they will need when they enter the workforce.

1. Through research, students can find and set up your own problems and solve them.
2. Students can read and understand technical literature, and prepare technical documents.
3. Students participate proactively in research discussions and present research results in an easy-to-understand manner.

Evaluation will be based on the content of the presentation and report (50%) and the level of understanding through discussion and questions (50%). (Individual evaluation criteria may be set for each laboratory.)

Japanese

Data Science/Engineering

Focus on topics related to deep learning, topic map databases, data mining, and text mining.
The research themes for FY2023 were as follows.
- Proposal of a method for discriminating between sentences generated by large-scale language models and human-generated sentences.
- Proposal of a method for estimating the location of emergency vehicles by sound
- Proposal of a method to support Japanese accent learning for international students
- A shape-robust trash classification method with limited data
- A proposal for an automatic offsides determination method.
- A GPT-based personal chatbot implementation based on prompting
- A method for analyzing emotions while watching a movie using eye tracking.
- A proposal for a hostile attack on speaker matching systems
- Proposal of a method for predicting moving demand using machine learning.

Specified as needed

Specified as needed

• Office hours differ for each academic advisor, so please follow the instructions at the first meeting.

Non-regionally-oriented course

• Course that cultivates an ability for utilizing knowledge
• Course that cultivates a basic interpersonal skills
• Course that cultivates a basic self-management skills
• Course that cultivates a basic problem-solving skills

Most classes are interactive

• 9.INDUSTRY, INNOVATION AND INFRASTRUCTURE