Search Colleges, Exams, Schools & more
Login
Professional Certificate in Machine Learning and Artificial Intelligence
Online
25 Weeks
3,779 POUND Quick Facts
| Medium Of Instructions | Mode Of Learning | Mode Of Delivery |
|---|---|---|
| English | Self Study, Virtual Classroom | Video and Text Based |
Courses and Certificate Fees
| Certificate Availability | Certificate Providing Authority |
|---|---|
| yes | Imperial College Business School, London |
The fees for the course is :
| Fees components | Amount |
| Course fees | £3,779 |
The Syllabus
- Understand the programme structure, core competencies and learning resources for the Professional Certificate in Machine Learning and Artificial Intelligence. Meet the four faculty members through an overview video, and explore the topics they will cover.
Refresh essential Python concepts through short videos. Learn to navigate EMCodE and Google Colab using step-by-step guides, practice exercises and troubleshooting tips.
- Apply linear algebra concepts — specifically vectors, matrices, transformations and eigenvalues — to solve mathematical problems in ML/AI.
- Analyse key calculus concepts — specifically derivatives, partial differentiation, chain rule and norms.
- Analyse key optimisation concepts, including gradient descent, local minimum and global minimum and learning rate.
- Apply optimisation techniques in Python.
- Examine the foundational mathematical concepts in ML and AI.
- Analyse the basics of ML, including key concepts such as variable classification, forecasting versus inference and the differences between ML and statistics.
- Examine the major dividing lines in the ML landscape, including supervised vs unsupervised learning, prediction vs classification, parametric vs non-parametric approaches and the ML process.
- Implement techniques to handle missing data in Python and learn the initial steps of the ML process.
- Assess whether ML is the right solution for specific organisational challenges and apply it to industry problems.
- Examine core probability principles and their practical applications.
- Differentiate between probability and statistics in data analysis.
- Apply probability models, including fair fair and biased coins, to real-world scenarios.
- Analyse independent and dependent variables and their impacts on calculations.
- Run Monte Carlo simulation to demonstrate probability principles.
- Implement Bayes' rule and total probability in decision-making contexts.
- Analyse probability distributions and their importance in ML.
- Analyse how statistics are used in real-world ML applications.
- Calculate measures such as mean, median and standard deviation and detect data outliers.
- Perform maximum likelihood estimation and analyse data relationships using regression and coefficients.
- Visualise data with scatterplots and interpret variable correlations.
- Evaluate statistical estimations using bootstrapping techniques.
- Examine how sample size impacts model accuracy and predictions.
- Evaluate different models and determine the most suitable one for specific data sets.
- Analyse the role of training, validation and test sets in model selection.
- Apply Laplace's rule of succession to compute probabilities.
- Assess the feasibility of ML for real-world problems through data analysis.
- Measure regression and classification performance using accuracy, sensitivity and specificity.
- Implement confusion matrices in Python to assess classification models.
- Use lift charts and ranking techniques to evaluate predictive models.
- Summarise model performance metrics for different types of output variables.
- Examine real-world applications of performance evaluation through competitions.
- Apply stratified sampling to balance class distributions in data sets.
- Use oversampling techniques to improve classification model performance.
- Evaluate predictive performance using k-fold cross-validation in Python.
- Combine predictive techniques to enhance model accuracy across industries
- Analyse KNN applications and calculate distance norms.
- Examine how predictors influence classifier performance.
- Apply normalisation techniques and differentiate categorical predictors.
- Evaluate decision boundaries and choose optimal k values for performance.
- Use KNN for classification and regression, optimising model performance.
- Evaluate decision tree models.
- Calculate information gain using entropy and the Gini index.
- Identify optimal data-splitting techniques to improve model accuracy.
- Build decision tree models in Python.
- Examine strategies to prevent organisational churn using decision trees.
- Determine optimal tree depth for accurate predictions.
- Enhance model accuracy using bagging, random forests and boosting techniques in Python.
- Evaluate real-world applications of tree-based models.
- Compare decision trees and KNN methods to choose the best approach for specific problems.
- Analyse model interpretability and fairness in ML.
- Identify when to use Naïve Bayes for decision-making.
- Apply Bayes's theorem and Naïve Bayes classifiers for predictions.
- Evaluate the strengths and limitations of the Naïve Bayes approach.
- Convert numerical data into categorical formats for improved model performance.
- Examine real-world applications of Naïve Bayes across industries.
- Use Python to fine-tune models and improve performance metrics.
- Analyse parameter tuning and its relationship to surrogate models.
- Balance exploration and exploitation in Gaussian processes.
- Determine the right time to stop tuning for optimal performance.
- Apply Bayesian optimisation to industry-specific challenges.
- Compare linear and logistic regression to understand their applications.
- Adjust parameters to optimise logistic regression models.
- Identify significant input variables using z-scores and p-values.
- Optimise logistic functions to enhance model performance.
- Evaluate the advantages and trade-offs of logistic regression compared with other methods.
- Select optimal hyperplanes to improve decision-making and classification accuracy.
- Examine the differences between hard-margin and soft-margin SVMs and apply them to data sets with outliers.
- Choose and implement the appropriate kernel type for different ML problems.
- Apply SVMs in Python for high-dimensional data and multi-class classification.
- Analyse real-world use cases of SVMs in organisational problem-solving.
- Analyse how neural networks approximate continuous functions to improve predictive performance.
- Develop effective neural network models for practical applications.
- Examine the fundamentals of gradient descent and its role in model optimisation.
- Apply backpropagation techniques to enhance network training efficiency.
- Select appropriate optimisation methods to improve network performance
- Analyse the five core building blocks of deep learning.
- Choose the right coding libraries and packages for deep learning projects.
- Construct neural networks and define effective decision boundaries.
- Assess the feasibility of deep learning for solving organisational challenges.
- Examine the biological inspiration and filtering operations behind CNNs.
- Analyse popular CNN architectures, including LeNet-5.
- Customise CNN models using PyTorch.
- Apply CNNs to real-world organisational problems and assess their performance.
- Analyse the role of hyperparameters in model performance.
- Evaluate different methods for tuning hyperparameters and their effectiveness.
- Apply hyperparameter tuning techniques to address specific organisational challenges.
- Examine emerging techniques and future applications in hyperparameter optimisation.
- Analyse the ability to understand language between humans and machines using historical and contemporary examples.
- Analyse the implications of different model sizes in relation to performance, accessibility and global Al development trends.
- Apply the concept of emergence to prompt engineering.
- Analyse how transformers work in relation to attention, parameter settings and model behaviour.
- Calculate hyperparameter sensitivity in the context of large language models.
- Analyse the impact of hyperparameter tuning on training time and scalability.
- Calculate the total attention comparisons in relation to model complexity.
- Analyse foundational LLM concepts, including emergence, transformer, hyperparameter tuning and attention.
- Explain how LLMs work, including their architecture, scale, training dynamics and emergent capabilities.
- Evaluate scaling strategies and hyperparameter tuning in relation to performance, cost and robustness.
- Examine the practical implications of emergence thresholds in relation to task-specific performance.
- Analyse the real-world value of few-shot learning in accelerating model training and deployment across different industries.
- Evaluate the risks of scaling LLMs in the context of emergence.
- Analyse the benefits of building a transformer from scratch in comparison to using prebuilt models.
- Analyse the foundational components of a transformer in relation to model structure.
- Refine a transformer in PyTorch based on changes to architecture.
- Analyse advanced LLM concepts in relation to scaling, emergence and transformer components.
- Examine the role of transparency in building trust, fairness and accountability.
- Identify bias in data and machine learning models.
- Apply documentation practices, including datasheets and model cards.
- Evaluate trade-offs between explainability, accuracy and fairness.
- Build interpretable models, including decision trees.
- Assess ethical implications alongside performance metrics.
- Differentiate between single data points and data clusters.
- Implement hierarchical clustering and k-means clustering in Python.
- Evaluate the strengths and limitations of clustering analysis.
- Apply clustering techniques to real-world organisational scenarios.
- Identify principal components that capture data variance.
- Determine the optimal number of components for effective dimensionality reduction.
- Apply principal component analysis (PCA) techniques using Python to real-world data sets.
- Explore practical applications of PCA in ML.
- Analyse the fundamentals of reinforcement learning and its real-world applications.
- Develop solutions for multi-armed bandit problems in Python.
- Apply Markov decision processes and Q-learning for optimal decision-making.
- Explore reinforcement learning applications across industries.
- Apply your machine learning skills to a realistic challenge by iteratively improving a model in a simulated black-box environment. This capstone runs throughout the programme and culminates in a portfolio-ready GitHub project that demonstrates your ability to tackle complex, real-world ML problems.