AI has become a big social issue as it spreads rapidly to science, industry, and even daily life. Chemical research based on AI using big data in chemistry has been reexamined. In this course, we will discuss the role of AI in modern chemistry and investigate the latest trends in this field. It aims to learn practical knowledge that can be used in actual chemical research field through theory and practice focused on deep learning. Student will learn how to link modern AI techniques to their own research, using both experimental and computational data. After successfully finishing this course, students would be confident in understanding and implementing AI models for chemical/molecular applications.

교과목 개요 : 계산화학, 특히 분자동역학 시뮬레이션 기법과 인공지능 기법의 기초를 배우고, 이를 활용한 여러 화학 분야의 최신 연구 데이터를 예시로 하여 화학연구에 계산화학 및 인공지능을 활용하는 법을 배운다.
교육 목표: 1) 기계학습과 인공신경망을 비롯한 인공지능 기법의 기초를 배운다.
2) 계산화학, 특히 분자동역학의 기초 개념과 시뮬레이션 활용기법을 배운다.
3) 계산화학과 실험화학의 최신 연구 데이터를 기반으로 1,2)에서 학습한 인공지능 및 계산화학 기법의 활용법을 실습을 통해 배운다.
3) 각자의 연구와 직접적으로 관련된 데이터를 사용하여 인공지능 기법을 연구에 활용하는 프로젝트를 수행함으로 연구활용능력을 배양한다.

CHEM211 or introductory knowledge on linear algebra/differential equation are recommended.

Each assignment (10%) * 7, Final project (30 %), Bonus (10 %)
No attendance score, but unapproved absence of more than 5 times will result in course failure. Not finishing final project will also result in F.

Bonus points will be given to those who actively participate in the class (ex. Questions). Either in class or after class, you may raise questions. A Q&A board is given through the PLMS site for the after-class participation.

The lecture will be given in English. Sometimes students will need to watch pre-recoded video lecture prior to coming class, and in class will be more of hands-on computational practices.

Note that the tentative schedule below will change as the course progresses.
More of molecular simulation contents will be included.
Week 01 : Introduction & math review
Week 02 : Machine learning fundamentals
Week 03 : Support vector machine (SVM)
- expressing chemical structure with SMILES
Week 04 : Monte Carlo method for data generation
- computational mutation studies / reaction scope generation

Lecture 12: Monte Carlo & Coupled HO

Week 05 : Deep learning & multilayer perceptron (MLP)
- learning structure-property relations from chemical database
Week 06 : Convolutional neural network (CNN)
- image processing, phase transition and drug discovery
Week 07 : Generating learnable chemical data – preparing for your final project

- Midterm week –

Week 08 : Molecular graph and Graph Neural Network (GNN)
- with applications in AI based synthesis
Week 09 : AI in computational quantum chemistry research
- with practical guide to DFT for non-computational chemists
Week 10 : AI in molecular dynamics simulations
- optimizing next-generation molecular models with ML
Week 11 : Variational Autoencoder (VAE)
- learning in latent space with examples in protein folding
Week 12 : Reinforcement Learning (RL)
- refining/reconstructing spectroscopic data w/ ML
Week 13 : Clustering and pattern recognition
- identifying nanostructures from low resolution TEM images

Week 14-15 : Student presentation for the results of their own term project

Chemical research areas to be covered : AI based synthesis, drug discovery, materials design,
refining spectroscopic data, optimizing reaction process, Monte Carlo methods,
image recognition and classification of nanostructures and phase transition,
learning structure-property relationship, DFT based mechanistic study,
optimizing molecular models and next generation force fields.

General : The lecture will be given in English. Sometimes students will need to watch pre-recoded video lecture prior to coming class, and in class will be more of hands-on computational practices.

Final project : During the last two weeks, lectures will be given by the students to give short presentations (7+3 min. for each student). Students are encouraged to pick a topic related to both the class and their own research interest. A set of topics would be provided after preliminary exam for the students who cannot find their own topic. The presentation must include a practice of AI technique in the selected topic.

Assignments : Assignments will be given (approximately) biweekly, and are the most important component of the class from a pedagogical point of view. Assignments will be due in class one week after distribution. You may discuss homework with other students, but must write up/code your answers independently.

You are encouraged to study together and to discuss information and concepts covered in lectures with other students. You can give "consulting" help to or receive "consulting" help from such students. However, this permissible cooperation should never involve one student having possession of a copy of all or part of the work done by someone else. Should copying occur, both the student who copied work from another student and the student who gave material to be copied will automatically receive a zero for the assignment. Penalty for violation of this Code can also be extended to include failure of the course and University disciplinary action.