Examination arrangement

Course content

This course assumes that you know close to nothing about Machine Learning (ML). Its goal is to give you the concept, the intuitions, and the tools you need to implement programs capable of learning from data. Throughout the course, we will cover several applied techniques, from the simple and the most common used (such as linear regression) to some of the Deep Learning techniques. In this course, you will learn how to implement your own toy versions of some algorithms and use actual frameworks in MATLAB and Python to work with benchmark problems and real-world datasets. Throughout the course, you will grow your own understanding of ML through concrete examples and a little bit of theory. In the classroom, you need to pick up your laptop, experiment with code examples provided in the book, and try your own code. Your participation, attendance, and ability to meet deadlines are crucially important to benefit from the course. The course contents are as follows, but not limited to:[1] Introduction[2] Supervised learning[3] Unsupervised learning and preprocessing[4] Representing data and engineering features[5] Model evaluation and improvement[6] Dimensionality reduction[7] Artificial neural networks and gradient descent[8] Convolutional neural networks[9] Recurrent neural networks[10] Training deep neural nets for images, video, text, and time-series

Learning outcome

Upon completion of the course, students will be expected to:• Student will: lean basic concepts of machine learning such as supervised and unsupervised learning, data preprocessing, learning from data, different deep learning models, transfer learning and finally model selection and evaluation. • Develop: students will learn how to develop a pipeline model that can be deployed to an endpoint using different machine learning platforms. Be able to design and implement various machine learning algorithms in a range of real-world applications. Have a good understanding of the fundamental issues and challenges of machine learning: data, model selection, model complexity, etc. • Utilize: student will learn how to utilize different machine learning platforms such as deep learning toolbox from MathWorks, and Python libraries such a keras, panda, tensrflow, etc.• Understand key elements of how to use machine learning in applications that require images, videos, text, timeseries, etc. Have an understanding of the strengths and weaknesses of many popular machine learning approaches. • Conduct research and apply tools and technologies in different areas such as HRI, image segmentation, sentiment analysis, text understanding and text summarization, object detection and tracking, tec.• Gain an understanding of the strengths and weaknesses of many popular machine learning approaches and models. Have an understanding of the underlying mathematical relationships within and across Machine Learning algorithms and the paradigms of supervised and un-supervised learning.• Develop an awareness and understand of how to reliably use machine learning models in different areas.

Learning methods and activities

Lectures for 4 hours along with exercises covering the entire course. Other supporting material (video, text, etc.) may be used.Mandatory assigments: 3 to be eligable for the oral exam

Compulsory assignments

• Assignments

Further on evaluation

Oral exam.

Re-sit exam is in May/June.

Recommended previous knowledge

the course does not presume or require any prior knowledge in machine learning. However, to understand the concepts presented and complete the exercise, we recommend that students meet the following prerequisites:• Mastery of intro-level of algebra: you should be comfortable with variables and coefficients, linear equations, graphs of functions. Being familiar with more advanced math concepts such derivatives is very helpful but not requires.• Proficiency in programming basics: it is highly recommended that you feel comfortable reading and writing Python code before starting this course. The programming exercise in the machine learning course uses Keras and pandas libraries.• It is expected that Matlab programmers will find no problem to switch to Python. Matlab users will need to install Deep Learning Toolbox™ from MathWorks.

Course materials

Reading list:Recommended texts, but not limited to, for a reading list:• A. Geron. Hands-On Machine Learning with SciKit-Learn & TensorFlow: concepts, tools, & techniques to build intelligent systems. 2017, O’Reilly.• P. Harrington. Machine Learning in Action. 2012, Manning.• J. Patterson and A. Gibson. Deep Learning: a practitioner’s approach. 2017, O’Reilly.• N. Buduma. Fundamentals of Deep Learning: designing next generation machine intelligence algorithms. 2017, O’Reilly.• D. Foste. Generative Deep Learning: teaching machines to paint, write, compose, and play. 2019, O’Reilly.

Credit reductions