| layout |
|---|
default |
In recent years we witnessed a huge development in machine learning, especially in deep learning which drives a new technological revolution. These models improve searches, apps, social media and open new doors in medicine, automation, self-driving cars, drones and almost all fields of science. In this introductory deep learning class students will learn about neural networks, objectives, optimization algorithms and different architectures. During the semester students will work on multiple projects: 2 smaller projects, where students try out different algorithms and architectures. To successfully complete the class, prior knowledge in Python (numpy, pandas, matplotlib) is required. During the course the students will learn about and will get comfortable with popular deep learning frameworks, mainly Keras and Tensorflow.
- Neptun: deeplea17m
- Language: Hungarian (slides and the materials in English)
- Start: 2020.02.14.
- Time: 12:45-13:45
- Location: 0.79 Jánossy Lajos room, North building
- Use issues on the course's GitHub page
- room 5.103 in North building
- Csabai István
- Olar Alex
- Pataki Bálint Ármin
- Bagoly Attila (former)
- Ribli Dezső (former)
- Second Kaggle challenge is live! Access via this link!
- from 6th March the lecture will start 15 mins earlier! (from 12:45 to 13:45)
- register to Kaggle
- please fill this form (to match Kaggle usernames and Neptun codes).
- First Kaggle challenge is available through this link. It is a limited competition, so you can download the data only via this link.
| week | topics | instructor | materials | date |
|---|---|---|---|---|
| 1 | Course introduction, technical details | Csabai István | intro, requirements | 2020.02.14. |
| 2 | Machine learning: ideas, concepts | Pataki Bálint Ármin | 2020.02.21. | |
| 3 | Machine learning model zoo & examples | Pataki Bálint Ármin | PDF, notebook | 2020.02.28. |
| 4 | Fully connected neural networks | Pataki Bálint Ármin | PDF, notebook | 2020.03.06. |
| 5 | Convolutional neural networks | Pataki Bálint Ármin | PDF, notebook, video | 2020.03.27. |
| 6 | Practical CNN, pre-trained models | Pataki Bálint Ármin | PDF, notebook_inference, notebook_finetune, video | 2020.04.03. |
| 7 | CNN visualization, self-supervised learning | Pataki Bálint Ármin | PDF, vis_notebook, video | 2020.04.17. |
| 8 | Modern CNN architectures | Olar Alex | PDF, notebook, video1, video2 | 2020.04.24. |
| 9 | Adversarial examples, GANs | Pataki Bálint Ármin | PDF, notebook, video | 2020.05.05. |
| 10 | Object detection, segmentation | Olar Alex | PDF, video | 2020.05.09. |
| 11 | Sequential models, recurrent neural networks | Pataki Bálint Ármin | PDF, wv_train, wv_check, video | 2020.05.15. |
- basic linear algebra
- basic probability and statistics
- Python (numpy, pandas, matplotlib)
During the semester there will be two Kaggle in-class challenges. Grades will be recieved by successful participation in them. For detailed description of the grading system, please visit the technical details slides.