Machine Learning Using Python Certification Course

• Probability and statistics
• Coordinate Geometry
• Linear Algebra
• Eingenvalues Eigenvectors and Eigendecomposition
• Introduction to Calculus

• Introduction to Statistics
• Understanding the Data
• Descriptive Statistics
• Data Visualization
• Probability
• Probability Distributions
• Sampling and Sampling Techniques
• Inferential Statistics
• Application of Inferential Statistics
• Relation between Variables
• Application of Statistics in Business
• Assisted Practice

Start this programme by understanding the course sections and the topics covered. This will help you be prepared for the upcoming sessions.

The course covers the basic concepts of machine learning, including its definition and different types. It also takes a deeper look at the machine learning pipeline, MLOps and AutoML, providing insights into deploying machine learning models at scale. In addition, students are introduced to the main Python packages for machine learning tasks, enabling them to use Python's robust ecosystem to develop machine learning solutions.

Topics:

• What is machine learning?
• Different types of machine learning
• Machine learning pipeline, MLOps and AutoML
• Introduction to Python packages for machine learning

The section on supervised learning explores its practical applications in different domains and is accompanied by discussions on its relevance and importance in real-world scenarios. Students engage in practical activities to prepare and shape data for supervised learning tasks, followed by discussions on overfitting and underfitting. In addition, practical exercises are offered to detect and avoid these problems, as well as insights into regularisation techniques to optimise model performance and reduce overfitting.

Topics:

• Supervised learning
• Applications of supervised learning
• Overfitting and underfitting
• Regularisation

This segment explores the basics of regression analysis, covering the definition and different types, including linear, logistic, polynomial, ridge and lasso regression. Discussions highlight the critical assumptions underlying linear regression and practical exercises provide hands-on experience in linear regression modelling. Participants will also engage in data exploration using techniques such as SMOTE oversampling and preparing, building and evaluating regression models to become proficient in regression analysis.

Topics:

• What is regression?
• Types of regression
• Linear regression
• Critical assumptions for linear regression
• Logistic regression
• Oversampling with SMOTE
• Polynomial regression
• Ridge regression
• Lasso regression

This section covers classification algorithms and their definitions, types and applications, and the choice of performance parameters. Participants are immersed in various classification techniques, such as Naive Bayes, Stochastic Gradient Descent, K-Nearest Neighbours, Decision Trees, Random Forest, Boruta and Support Vector Machines, through discussions and guided exercises. Key concepts such as Cohen's Kappa are also discussed, followed by knowledge checks to reinforce understanding.

Topics:

• What are classification algorithms?
• Different types of classification
• Types of applications and choice of performance parameters
• Naive Bayes
• Stochastic gradient descent
• K-neighbour populations
• Decision tree Random Forest
• Boruta
• Support vector machine
• Cohen's mantle

This segment introduces students to unsupervised algorithms, covering their types, applications, and performance parameters. Participants engage in hands-on activities such as visualizing output and applying techniques such as hierarchical clustering, K-Means clustering, and the K-Medoids algorithm. In addition, they explore anomaly detection methods and dimensionality reduction techniques such as Principal Component Analysis (PCA), Singular Value Decomposition, and Independent Component Analysis. Practical applications of these algorithms are demonstrated through guided exercises, enhancing students' understanding of unsupervised learning concepts.

Topics covered:

• Unsupervised algorithms
• Different types of unsupervised algorithms
• When to use unsupervised algorithms?
• Parameters for performance
• Types of clustering
• K-Means clustering
• K-Medoids algorithm
• Outliers
• Detection of outliers
• Principal component analysis
• Correspondence analysis and multiple correspondence analysis (MCA)
• Singular value decomposition
• Independent component analysis
• Balanced iterative reduction and clustering using hierarchies (BIRCH)

This section covers classification algorithms and their definitions, types and applications, and the choice of performance parameters. Participants are immersed in various classification techniques, such as Naive Bayes, Stochastic Gradient Descent, K-Nearest Neighbours, Decision Trees, Random Forest, Boruta and Support Vector Machines, through discussions and guided exercises. Key concepts such as Cohen's Kappa are also discussed, followed by knowledge checks to reinforce understanding.

Topics:

• What are classification algorithms?
• Different types of classification
• Types of applications and choice of performance parameters
• Naive Bayes
• Stochastic gradient descent
• K-neighbour populations
• Decision tree Random Forest
• Boruta
• Support vector machine
• Cohen's mantle

This module provides a comprehensive overview of recommendation engines and explores their underlying principles and mechanisms. Participants are immersed in various use cases and examples of recommendation systems and gain insight into their design and implementation. Through practical exercises, participants apply collaborative filtering techniques, including memory-based modelling, object-based and user-based filtering, and model-based collaborative filtering. In addition, they explore dimensionality reduction, matrix factorisation methods and accuracy matrices in machine learning to evaluate and optimise the performance of recommendation engines.

Topics:

• How do recommendation machines work?
• Use cases for recommendation machines
• Examples of recommendation systems and how they are designed ¨
• Using PyTorch to build a recommendation engine.

At the end of the course, you will do two projects. You will apply all you have learned and gain practical experience with your new knowledge.

• Project 1: Employee turnover analysis - Create ML programmes to predict employee turnover, including data quality checks, EDA, clustering, etc., and propose employee retention strategies based on probability scores.
• Project 2: Segmentation of songs - Conduct exploratory data analysis and cluster analysis to create cohorts of songs.