Data Science

---

Pathway

We’ll kick off by getting to grips with the fundamentals of Python. As developers you will already know how to code, but not necessarily in Python or using the stats libraries and functions required for data science. Here you will learn why Python is the go-to data language.

Now we are ready to start looking at data science. What is it? What can you use it for and why? We’ll explore all these questions and more as we begin the journey into data.

Time to travel back to school and get a refresher in foundational maths, histograms, mean, median and mode, standard deviation and more. All the basic statistics you will need.

Moving on from foundational maths, we will now look at basic probability theory, covering combinatorics, bayesian inference, and distributions.

Next we will look at inferential statistics and hypothesis testing, where we will begin reaching conclusions that extend beyond the data available.

What software is in the toolkit of a data scientist? In this session we will look at a range of tools used throughout the industry, including Jupyter Notebook, NumPy and pandas, as well as how we can better use Visual Studio Code with the right extensions for your project.

Now that we have learnt about different ways of analysing our data, we will start to visualise it. Using tools like Matplotlib, seaborn and pandas, we will start to display our data in different ways.

This session is part one of understanding how to load data into Python, read it, and clean it.

Following on from the previous class, we will now explore combining dataframes and other more complex methods of working with and cleaning data.

Time to step it up a bit. We’ll now look at more advanced data methods, what they do, and how we can use them to create meaningful outcomes.

The really interesting work begins here. With a range of real data sets we are now ready to start cleaning and analysing data to begin making predictions.

Now that you have analysed some real data, you will present your work to the team and open it up for discussion.

The halfway point! With data science under your belt, we will now move on to machine learning. We’ll start with basic definitions and examples of what you can do.

We’ll now start looking at a range of regression techniques (some simple and some complex), used to analyse relationships between variables in your data. Techniques such as multiple linear regression, Support Vector Regression (SVR), decision trees and random forests will all come into play.

Some data needs to be put into categories, or classifications, in order to better analyse it. Logistic regression, Support Vector Machine (SVM), and K Nearest Neighbour (KNN) are just some of the classification techniques we will cover.

Working with time series data can be tricky – now we will look at how to analyse this type of data and work with Hadoop Distributed File System (HDFS).

There are lots of concepts to learn when working with machine learning models. In this module we will learn about concepts including over/underfitting, bias/variance, confusion matrices, and precision vs accuracy.

Now that we know how to interpret different models, we will learn about optimising those models using techniques such as hyperparameter tuning, data augmentation, grid search and k-fold validation.

We are now ready to start a group project. Working with real data you will clean the data, select a model, and optimise it.

What is deep learning? In this module we will discover what it is, what we can use it for and how it works. We will also discuss different libraries (sklearn, pytorch, keras, tensorflow) and when to use them.

Following on from our initial discussion of deep learning and the libraries involved, we will use keras, a TensorFlow library, to create a neural network.

Now that we have looked at creating neural networks, we will explore the maths behind it, as well as the processes involved and what we can use keras for, including transfer learning.

Having learnt all about machine learning, we will now learn how to integrate a machine learning model into a website.

In our final module, you will present your machine learning projects. We will discuss them, and then finish the course by talking over next steps for your new career in data science.