Machine learning influences modern life in many different avenues and is silently revolutionising the way we live and work. We can see the influence of machine-learning algorithms in social media, web search engines, mobile device spell checkers and self-driving cars. This course will give you an introduction to machine learning using the Python programming language and the TensorFlowTM programming toolkit from Google. No programming background is assumed, however if you want to take this course, you should be familiar with using computers.
This course is taught by scientists using machine learning for data analysis at CERN’s Large Hadron Collider and will allow you to work on practical examples from both general and physics-based problems. Examples will be drawn from a variety of problems in order to allow you to build up an understanding of the tools and how to use them. This will prepare you for a mini-project analysing data from a particle physics experiment to complement the examples encountered earlier in the course.
This is a practical course that provides you with an introduction to the concepts of machine learning and the application of algorithms to several types of available data samples. In order to achieve this, you will be introduced to the Python programming language and key concepts related to the TensorFlowTM programming toolkit. You will learn how to train machine-learning algorithms and evaluate their performance on image data and scientific data from the Large Hadron Collider. We will develop your programming skills so that you can explore the potential benefits of deep-learning algorithms.
Teaching and learning
You will be taught through a combination of lectures, laboratory work, and workshops.
You will learn/develop:
- basic commands in Python and learn how to manipulate data using this programming language
- how to use TensorFlowTM tools to optimise neural networks and convolutional neural networks as examples of machine-learning algorithms
- a comprehension of machine-learning algorithms and their use.
You will develop/be able to:
- understand the principles of optimisation algorithms and the role of activation functions in neural networks
- understand the concept of overtraining of hyperparameters for a machine-learning algorithm, and how that can be spotted using data samples
- understand the concept of the Receiver Operating Characteristic (ROC) curve and how the area under this curve can be used to select models based on the ability to separate signal from background
- demonstrate information expertise through the portfolio of work that you will build during this course, and the application of that portfolio of skills to problem solving
- demonstrate a rounded intellectual development in all aspects of this course, including self-study, directed reading, in-session quizzes to test your incremental assimilation of knowledge and the final critical presentation of what you have learned and achieved during the course
- improve your research capacity via the application of core principles on machine learning to example data sets. This will allow the critical analysis of data in terms of specific problems using modern techniques
- communicate clearly via the oral presentation component, where you will give a five-minute presentation on what you have learned during the course (including the main results you have obtained) and will respond to questions on your presentation.
To join our Summer School, you should have completed a minimum of two semesters’ study at your home institution.
We welcome Summer School students from around the world. We accept a range of qualifications:
- if your home institution uses the four-point Grade Point Average (GPA) scale, we usually require a 3.0 GPA
- if your home institution uses the letter scale, you will need to have a B+
We welcome international qualifications and we consider every application individually on its academic merit.
English language requirements
All of our courses are taught and assessed in English. If English isn’t your first language, you must meet one of the following English Language requirements in order to join the QMUL Summer School:
- If you hold a degree from a majority English speaking country plus Canada you may use this degree to satisfy the English language requirements for entry, provided the degree was completed no more than 5 years before the start date of the course to which you are applying.
- IELTS, 7 overall or higher
- TOEFL Internet Based Test we require a minimum of 100 (L22; S25; R24; W27)
- PTE Academic 68
- Cambridge Certificate in Advanced English 185 70- grade C (old marking system)
Esta escola oferece programas em:
Última actualização January 18, 2018