Welcome to MLP!
Dr. David Elliott
Machine Learning in Python
What is Machine Learning?
Machine learning (ML) aims to automatically detect patterns in data to predict future outcomes of interests and make decisions.
Aims
We aim to teach some of the fundamentals of ML, not viewing it just as a black box but understanding:
- Theory, ideas, advantages, and limitations of the methods.
- How to select, compare, and evaluate different methods.
- Implementation in practice with Python.
We will introduce a large number of methods...
We will introduce a large number of methods...
...from the simplest and most commonly used (e.g. Linear Regression)...
We will introduce a large number of methods...
...from the simplest and most commonly used (e.g. Linear Regression)...
...to more complex ensemble methods such as Deep Learning.
In the lectures we will learn the theoretical foundations of these approaches, and apply this knowledge to weekly exercises.
In the workshops, rather than implementing our own toy versions of algorithms on toy data, we will use production-ready Python frameworks on "real" research data.
In the workshops, rather than implementing our own toy versions of algorithms on toy data, we will use production-ready Python frameworks on "real" research data.
Scikit-Learn
In the workshops, rather than implementing our own toy versions of algorithms on toy data, we will use production-ready Python frameworks on "real" research data.
Scikit-LearnKeras
Software
We will be using Jupyter Notebooks to write code and present our work.
Course FAQ
Q - What background experience does this course assume?
A - It is assumed students have prior experience with Python and understand the basics of...
- data types and structures (e.g.
numpyandpandas), - loops and vectorisation,
- data visualisation (e.g.
matplotliband/orseaborn).
Also some reasonable understanding of undergraduate-level maths is required.
(e.g. calculus, linear algebra, probabilities, statistics)
Q: Why Python?
A - It is a powerful yet simple language. It has become the most popular programming language for data science because it can allow us to quickly try ideas and put concepts into action.
One link to rule them all...
... where you can find everything except your course marks!
https://bit.ly/mlp-s2-22