Deep learning for science students
Animations of Yonsei's Eagle Statue generated by a GAN algorithm.
Note: the course was postponed for personal reasons.
Jump to: calendar
Course Description
In recent years, scientist have found applications of Machine Learning/Deep Learning in areas as diverse as pure mathematics (knot theory) or medicine (protein structure prediction). The industry has also benefit from the use of Deep Learning to automatize their process and create new products. In this class, our main goal is to provide the tools that students (math or science students with strong analytical skills but not much programming experience) require to be able to run their own experiments.
Requirements: Linear algebra and multilinear calculus.
Bibliography
You can find this books for free online:How to Think Like a Computer Scientist: Interactive Edition
Deep learning/Goodfellow, Ian/MIT Press/2017
Pattern Recognition and Machine Learning
You can find this books at the library:Deep Learning with Pytorch, First Edition
Deep Learning with Python, Second Edition
Time & Place
Lecture: Monday, Wednesday, Friday 10:00AM - 10:50PM Room 262 Yonsei's math department
Instructors, Office Hours, Online help
TA TBD
Policies
Grading
100pt for class project
100pt for group project
100pt for homeworks / presentations
Late Policy
If needed, the deadline of an assignment can be extended up to one week after the initial deadline with no evidence required from the student. Just send an email requestig the extension.
Cheating & Academic Dishonesty
If you have a question about whether something may be considered cheating, ask, prior to submitting your work.
Attendance
Attendance is not graded.
(Tentative) Course Calendar
주 |
기간 |
수업내용 |
교재범위,과제물 |
비고 |
March/2-March3 |
Introduction |
What is Machine Learning? Python review: Genetic Algorithms. Jupyter Notebooks |
|
|
March/7-March11 |
Data Oriented Machine Learning. |
The data pipeline. Neural networks. Pytorch vs Keras. Tensorboard. |
|
Course add and drop on March/8. March/9 Holiday. |
March/14-March18 |
Data Oriented Machine Learning. |
Optimization. Back propagation. Lightning. |
|
|
March/21-March25 |
Data Oriented Machine Learning. |
Regularization. Double descent. Testing. |
|
Friendly competition between Yonsei University Math Department and ESFM (Escuela Superior de Física y Matemáticas), Mexico |
March/28-April1 |
Deep learning for images. |
CNN. Cloud computing.
|
|
|
April/4-April/8 |
Deep learning for images. |
CNN. Pooling. Docker. |
|
|
April/11-April/15 |
Class project plan. |
|
|
|
April/18-April/22 |
Class project. |
|
|
Midterm examination period April/20-april/26 |
April/25-April/29 |
Class project report. Social Bias. |
Social biases and technology. Multiprocessing library. |
|
Midterm examination period April/20-april/26 |
May/2-May/6 |
Generative deep learning. |
GANS. Pix2pix. Profiling code. |
|
May/5 Holiday |
May/9-May/13 |
Deep learning for text. |
Recurrent NN, LSTM. Gradio vs Streamlit, huggingface, autoML. |
|
|
May/16-May/20 |
Deep learning for text. |
N-grams, Bag of words. Final group project check. |
|
|
May/23-May/27 |
Transformers. |
BERT and GPT. Final group project check. |
|
|
May/30-June/3 |
Deep learning for time series. |
Final group project check. Text to Image: VQGaN+CLIP, Disco Diffusion, DALL-E2, Google Imagen, Parti, Craiyon, Cogview2 |
|
June/1 Holiday |
June/6-June/10 |
Final group project check. |
Final group project check. |
|
June/6 Holiday June/8—June/14 supplementary lessons |
June/13-June/17 |
Final group project submission. |
|
|
June/8—June/14 supplementary lessons June/15-June/21 final examination period. |
June/20-June/24 |
|
|
|
June/22 Summer Vacation. |
|
|
|
|
|
- AI and the Everything in the Whole Wide World Benchmark
- Descending through a Crowded Valley — Benchmarking Deep Learning Optimizers
- WANLI: Worker and AI Collaboration for Natural Language Inference Dataset Creation
- Learning to Route by Task for Efficient Inference
- Why Does Deep and Cheap Learning Work So Well?
- Model Evaluation, Model Selection, and Algorithm Selection in Machine Learning
- Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets
- A ConvNet for the 2020s
- Collective Intelligence for Deep Learning: A Survey of Recent Developments
- The Annotated Transformer
- EvilModel: Hiding Malware Inside of Neural Network Models
- Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints
- Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
- Self-Diagnosis and Self-Debiasing: A Proposal for Reducing Corpus-Based Bias in NLP
- Towards Fairness in Visual Recognition: Effective Strategies for Bias Mitigation