Machine Learning with Python
Learn the foundations of algorithms for predictive analytics, and how to apply them to your data using Python.
About the course
This course helps analysts, researchers, BI experts and developers become familiar with building machine learning solutions using tools from the Python ecosystem.
Using a mix of presentation and interactive examples, the course compares supervised and unsupervised learning, and gives an overview of the core algorithms for classification, clustering, regression and dimensionality reduction. It also covers the basics of neural networks and recent developments in deep learning.
What you will learn
- Framing a business application as a machine learning task
- The role of labelled data, data cleaning and data transformation in machine learning systems
- Feature engineering techniques to extract useful attributes from your data
- Implementing supervised and unsupervised learning techniques in Python
- Evaluating the quality of your models
Syllabus
Overview
- What is artificial intelligence, and what's behind the hype?
- Data science vs data mining vs machine learning
- Machine learning problems and applications
- Python environment set-up with Anaconda
Machine learning concepts
- Learning and prediction
- Feature engineering
- Training data and test data
- Cross-validation
- Underfitting and overfitting
Supervised learning
- Classification: predicting a label
- Algorithms for classification: k-nearest neighbours, support vector machines, naive Bayes
- Regression: predicting a quantity
- Algorithms for regression: linear regression
Unsupervised learning
- Clustering: grouping similar items
- Algorithms for clustering: k-means, hierarchical clustering, DBSCAN
- Dimensionality reduction
- Algorithms for dimensionality reduction: principal component analysis
Evaluating machine learning algorithms
- Evaluation metrics for machine learning
- Planning an evaluation campaign on your data
Deep learning and neural networks
- Introduction to artificial neural networks
- Neural network concepts
- Neural network types
- Gradient descent
- Back-propagation
- Activation functions
- Loss functions
- Hyper-parameters
- Neural networks in the wild: examples of successful applications
- Deep network architectures