As data plays a larger part in every organisation, especially digital ones, it becomes more and more necessary to understand where data comes from, how it's stored, visualised, shared and manipulated.

‍Module 1 - Data collection
    Introduction to data pipelines
    Introduction to APIs
    HTTP requests
    Web scraping using requests and Selenium

Module 2 - Data formats & the Pandas Python library
    Important file formats
    Intro to Pandas
    More Pandas
‍
Module 3 - Data cleaning & exploratory data analysis (EDA)
    Reasons for and approaches to data cleaning
    Descriptive statistics
    Data visualisation
    Handling missing data
‍
Module 4 - SQL
    SQL basics
    SQL Join operations
    SQL Aggregations
    SQL Subqueries
    Creating databases and importing data
    Python SQLAlchemy

Module 5 - Extract, Transform, Load (ETL)
    ETL using Pandas, SQLAlchemy
    MapReduce
‍
Module 6 - Intro to cloud
    Intro to Amazon Web Services (AWS)
    AWS RDS
    The AWS CLI & Boto3
   AWS S3

This unit covers a comprehensive overview of data science. We'll start by learning how to visualise and explore data using industry standard tools. Modelling data is a key part of data science, and a key tool for doing this is machine learning (ML).
‍
Module 1 - Data Science Basics
    Statistical modelling
    Hypothesis testing
    Multi-collinearity
    Influential points

Module 2 - Introduction to machine learning
    Intro to ML & when (NOT) to use
    Scikit-learn
    Validation and testing
    Linear regression
    Logistic regression
    Multiclass classification
    Hyperparameters, grid search and K-fold cross validation
    Feature selection and feature engineering
‍
Module 3 - Theory
    Evaluation metrics
    Bias & variance, underfitting and overfittting
    Regularisation
    Maximum Likelihood Estimation (MLE)
‍
Module 4 - More supervised models
    K-nearest neighbours
    Decision trees
    Random forests
    Gradient boosting

Module 5 - Unsupervised models
    K-means clustering
    Dimensionality reduction using principal component analysis (PCA) & T-SNE
    DBSCAN

In this unit, we introduce deep learning models, a class of machine learning models that are able to represent much more complex relationships in data. This means they can reach state of the art results on tasks that involve processing images, audio, video and other complex data.

Module 1 - PyTorch
    Automatic differentiation
    PyTorch Datasets and DataLoaders
    Making custom datasets

Module 2 - Neural networks
    Neural networks
    Dropout
    Batch Normalisation
    Optimisation for deep learning
    Convolutional Neural Networks (CNNs)
    ResNets

Module 3 - Practical tips
    Architecture tips, data augmentation & debugging tips
    Pre-trained models
    Transfer learning
    Hardware acceleration (GPUs & TPUs)

Module 4 - Advanced Architectures
    Attention
    Self-attention
    Transformers

Module 5 - Applications
    Churn modelling
    Content based recommendation systems
    Collaborative filtering
    Lead scoring

In the ML Engineering part of this unit, we'll focus on the engineering required to put models into the real world. Then in the MLOps part, we'll focus on the tools which are becoming industry standards for ensuring that your AI systems scale up, are fault tolerant, and are well documented.

‍Module 1 - Scaling experiments in the cloud
    EC2, AMIs & SSH
    Experiment tracking
    Ray Tune
‍
Module 2 - Managed solutions
    AWS Sagemaker
    Architecting an end-to-end ML product on AWS
‍
Module 3 - Intro to MLOps
    MLOps overview
    ETL, feature stores & data versioning
‍
Module 4 - Custom solutions
    Docker
    Flask
    MLFlow
    TorchServe
    Prometheus
    Grafana
‍
Module 5 - Fault-tolerant deployment & autoscaling
    CI/CD with GitHub actions
    Kubernetes
    AWS EKS
    Kubeflow
    AWS Cloudformation