Welcome to you all !
This GitHub Home page contains all the materials for the course Machine Learning and Finance 2022-2023 at Imperial College Business School.
- The scripts are written as Jupyter notebooks and run directly in Google Colab.
- If you prefer to run the scripts on your machine, please follow the instructions in the following link: Installation instruction
- Basic knowledge in probability theory and calculus.
- Proficiency in some programming language, preferably Python.
| Date | Stream | Start | End | Lectures topics | Lectures | Quiz | Quiz Solution | Programming Session | Optional Reading |
|---|---|---|---|---|---|---|---|---|---|
| 04/17/23 | Both | 10:00 | 12:00 | Optional Session 1: Python Refresher | Practical Implementation | No quiz | No quiz | Code_Python Solution_Python |
|
| 04/18/23 | Stream2 Stream1 |
14:00 18:00 |
17:00 21:00 |
Fundamentals of Machine Learning | Lecture_1 | Background Slide8 Slide23 Quiz1 link Quiz1 pdf |
Quiz1_Solution | Programming_Session_1 Solution_Programming_Session_1 |
Optional_reading |
| 04/21/23 | Stream2 Stream1 |
09:00 13:00 |
12:00 16:00 |
Supervised Learning Algorithms | Lecture_2 | Slide30 Slide41 Slide47 Quiz2 link Quiz2 pdf |
Quiz2_Solution | Programming_Session_2 Solution_Programming_Session_2 |
|
| 04/24/23 | Stream1 Stream2 |
13:00 17:00 |
16:00 20:00 |
Practical Implementation : Credit risk dataset | Practical Implementation | Correct_Section Poll |
No quiz | Programming_Session_3 Solution_Programming_Session_3 |
|
| 04/25/23 | Stream2 Stream1 |
14:00 18:00 |
17:00 21:00 |
Introduction to Neural Networks. | Lecture_4 | Quiz4 link Quiz4 pdf |
Quiz4_Solution | Programming_Session_4 Solution_Programming_Session_4 |
Optional_reading |
| 04/28/23 | Stream2 Stream1 |
09:00 13:00 |
12:00 16:00 |
Introduction to Unsupervised Learning: Creating word vectors using the GloVe approach. | Practical Implementation | Introduction GloVe_Quiz Word2vec_Preprocessing Word2vec_Creating_Training_Data Word2vec_Training_process |
GloVe_Quiz_Solution Word2vec_Preprocessing_Solution Word2vec_Creating_Training_Data_Solution Word2vec_Training_process_Solution |
Programming_Session_5 Solution_Programming_Session_5 |
Optional_reading |
| 05/02/23 | Stream2 Stream1 |
14:00 18:00 |
17:00 21:00 |
Neural Networks for sequences. | Lecture_6 | Quiz6_link Quiz6 pdf |
Quiz6_Solution | Programming_Session_6 Solution_Programming_Session_6 |
|
| 05/04/23 | Both | 10:00 | 11:00 | Optional Session 2: Finishing Programming Session 3 | Practical Implementation | ||||
| 05/05/23 | Stream2 Stream1 |
09:00 13:00 |
12:00 16:00 |
Practical Implementation: Sentiment Analysis | Practical Implementation | Correct_Section Poll |
Programming_Session_7 Solution_Programming_Session_7 |
||
| 05/09/23 | Stream2 Stream1 |
14:00 18:00 |
17:00 21:00 |
Attention mechanisms and Transformers | Lecture_8 | RNN_Applications Alignment Attention_Weights Self_Attention |
Quiz8_RNN_Applications_Solution Quiz8_Alignment_Solution Quiz8_Attention_Weights_Solution Quiz8_Self_Attention_Solution |
Optional_reading | |
| 05/12/23 | Stream2 Stream1 |
09:00 13:00 |
12:00 16:00 |
Dummy exam and revision elements | Review | Review Stream 1 Stream 2 |
|||
| 05/15/23 | Both | 09:00 | 11:00 | Optional Session 3: Coding the Transformer architecture | Transformer | Mock Exam | Solution_2022 | Optional Session | Optional_reading |
| Session | Review pdf |
|---|---|
| Session 5 | Review 1 |
| Optional Session 2 | Hyperparameters Optimization |
| Session 7 | Review 2 |
| Session 9 | Review 3 Stream 1 Review 3 Stream 2 |
| Optional Session | Optional Session |
The module is structured around 9 sessions of 3 hours each. The sessions are comprised of lectures and practical implementation sessions. Students will be expected to devote an equivalent amount of learning time outside of class, in private and group study of module material. Some of the teaching format will employ Python.
The module will introduce the main subareas of Machine Learning in order to tackle various problem tasks. It is practicularly focused on a deeper understanding of sequence modeling using neural networks and attention mechanisms.
The objectives of this module are:
- Develop knowledge on the roadmap for building machine learning systems.
- Get familiar with traditional Machine Learning algorithms and more advanced techniques including Deep Learning.
- Get a good understanding of the basic concepts of Supervised Learning, Unsupervised Learning and Sequence Models.
- Develop skills to process sequential data, especially in the context of Natural Language Processing.
- Practice supervised learning by predicting loan default risk.
- Practice sentiment analysis with state-of-the-art algorithms for sequential data.
- Coursework : 50%
- Exam : 50%
| Assignment | Type | Weighting | Date Released to students | Date Due |
|---|---|---|---|---|
| Coursework | Group project | 50 % | 05/15/23 | 05/30/23 |
| Year | Coursework | Exam |
|---|---|---|
| 2023 | ||
| 2022 | Coursework_2022 Solution_2022 |
Exam_2022 Solution_2022 |
| 2021 | Coursework_2021 Solution_2021 |
Exam_2021 Solution_2021 |
| 2020 | Coursework_2020 Solution_2020 |
Exam_2020 Solution_2020 |
Please feel free to contact us if you have any questions or require further information at: [email protected]