Welcome back

Welcome to the second week of Applied Data Science. Last week we learned about fundamental data science topics, including how to explore the data with basic data analytics tools and how to pre-process the data by preparing and cleaning it to remove redundant and erroneous data.

We also explored basic visualisation and modelling techniques to identify promising patterns to build a model to distinguish landscape from portrait paintings using a dataset from the Tate Museum. Finally, we discussed what model evaluation is and why it is crucial to verify the performance of a model.

What you will be learning this week

This week we introduce the concepts of dimensionality in a dataset and the value of understanding what the most important characteristics of a dataset are. This will help us identify key features, normalise our dataset and hopefully be able to explore the knowledge it contains. We will discuss how to make decisions on feature selection, which of these characteristics are essential and how to effectively select the most prominent features to reduce noise and improve the performance of our model.

Introducing this basic concept will help us move forward with the introduction of the popular concept of machine learning, and will enable us to demystify machine learning by running through some simple examples using Python and modern machine learning libraries. We will then discuss the fundamental theory behind neural networks, one of the most popular machine learning algorithmic frameworks, and introduce the perceptron – which is the simplest neural network.

Your task

What are you most looking forward to learning about this week and why?

Please share your thoughts with your fellow learners in the comments.