2 Credits

In this course, the students will learn how to oversee, protect and make efficient use of data for various organizational settings. The organizational environments include government departments and offices, educational institutes, corporates etc. The courses focus on data management aspects with respect to organizations and data governance. The core topics of this module are:

Data Management vs Data Governance: Overview of data management, Characteristics of organization data, Data governance as a concept, Importance of data governance, analyzing the relationship with data management and data governance; Data Architecture: Models, policies, rules and standards that govern how data is collected, stored, arranged and integrated and used in systems/ organizations; Data Modeling and Design, Data Storage and Operations, Data Integration and Interoperability; Data Security: Technology-based methods: Software and Hardware methods, disk encryption, backups, data masking and erasure, Laws and Standards: Sri Lankan laws on data security; Data Maintenance, Metadata, DataQuality

2 Credits

This is an introductory course in data science where the student will learn the fundamental aspects of the discipline including the definitions, process and sub-disciplines (e.g. Business Analytics). In addition, the course discusses the state of the art applications in data science and business analytics and contemporary application areas and trends. The core focuses include:

Data: Definition of data, Characteristics of small data, massive data and big data, Structured vs Unstructured data; Data Science: What is Data Science? Big data and data science, Skills for data science, Importance of business analytics, Fundamental building blocks of data science; Data Wrangling: Various data sources and types, Aspects of big data, Data integration and processing techniques, Data preparation and integration techniques for business analytics; Overview of Data Modeling and Analysis Techniques: Supervised and unsupervised models, Associated rule mining, Exploratory techniques; Applications of Data Science as Case Studies

3 Credits

Programming and the ability to manipulate data using programming languages and relevant packages is an essential skill for any data scientist. This course introduces the data programming languages and associated essential third-party tools which enables the student to gain programming practice on developing solutions to business analytics problems. The main programming focus includes:

R Programming: Introduction to R Programming Language and R studio, Functions in R, Control Flow and Loops, Vectors and Metrics in R, Reading and Writing Data, Integration with Databases, Simulation and Profiling, Model fitting, Graphics packages in R; Python Programming: Introducing to Python Language, Lists, Functions and Packages, Control Flow and Loops, Scientific Packages (NumPy, Pandas, Mathplotlib etc.), git integration, Jupiter Notebooks; Contemporary Data Science Tools and Packages

3 Credits

Machine learning and pattern recognition belong to two core areas of contemporary data science. This course introduces the fundamental learning paradigms, pattern recognition algorithms and how this knowledge can be used to model business problems as a machine learning task. The core content of the course includes:

Fundamentals of Machines Learning and Pattern Recognition: Types of machine learning problem, Overview of machine learning algorithms, Application of machine learning and pattern recognition in business environments; Features: Feature extraction and feature selection methods; Classification: Types of classification, Bayesian decision theory and pattern classifications, Naïve Bayes classification, Bayesian Belief Networks, Tree-structured classification; Supervised Learning: Perceptron learning, Backpropagation learning and its variants, Random forests, Support vector machines; Unsupervised Learning: Hebbian learning, K-means concepts and its variants, Self-Organizing Maps and Neural Gas Algorithm; Combining Classifiers: Classifier fusion techniques, Bagging and boosting techniques.

3 Credits

This course introduces the concepts of data warehousing and data mining and their applicability to the modern-day data science landscape. Students will be introduced to critical components of a data warehouse architecture and to use these components to design data warehouses to support and provide efficient and practical solutions to business problems. The core content of the course includes:

Data Warehousing: Introduction to data warehousing, Data warehouse architectures, Multidimensional data model, Concept hierarchies, Attribute generalizations and relevance, Online analytical process, Business intelligence tools and visualizations; Data Mining: Fundamentals of data mining, the Knowledge extraction process, Data and data exploration, Frequent itemsets mining, Association rule mining, Classification and cluster analysis, Web mining.

3 Credits

It is said that a picture says a thousand words. Similarly, a properly done visualization of real-world data can sometimes solve the problem without any more inferences or modelling. Hence, this course presents the techniques of information visualization that enables the student to create efficient and impactful visualizations. The core content of the course includes:

General Concepts: Purpose of visualization, what is visualization? Functions of visualizations, why do we create visualizations? Visualization Design, Organization and Evaluation: Types of data in the visualization, Bertins’ levels of organization, Mackinlays’ Criteria, Perception, gestalt, Interaction methods, Layouts, Deconstruction visualization, Aspects of Color in Visualizations, Animations, Evaluating visualizations; Exploratory Data Analysis: Data wrangling, Types of models, Types of Variables, Statistical Model,
Hypotheses testing using graphical data, Visual Analytics

It is said that a picture says a thousand words. Similarly, a properly done visualization of real-world data can sometimes solve the problem without any more inferences or modelling. Hence, this course presents the techniques of information visualization that enables the student to create efficient and impactful visualizations. The core content of the course includes:

General Concepts: Purpose of visualization, what is visualization? Functions of visualizations, why do we create visualizations?; Visualization Design, Organization and Evaluation: Types of data in the visualization, Bertins’ levels of organization, Mackinlays’ Criteria, Perception, gestalt, Interaction methods, Layouts, Deconstruction visualization, Aspects of Color in Visualizations, Animations, Evaluating visualizations; Exploratory Data Analysis: Data wrangling, Types of models, Types of Variables, Statistical Model, Hypotheses testing using graphical data, Visual Analytics

3 Credits

This course will introduce to the students the core principles and challenges associated with predictive modelling and analytics of modern-day data. Modelling will discuss the widely used techniques and tools and their applicability towards various kinds of data problems. The core topics include:

An Introduction to Predictive Analytics: The predictive analytics process, different types of models (predictive, descriptive and decision), Application of predictive analytics, Mathematical basics for predictive analytics; Data: Types of data in the business domain, cleaning up data and data wrangling, Dealing with missing values and outliers, Adding and removing variables.; Regression Models: Linear regression, Nonlinear regression, Logistic regression, Multiple regression, Model diagnostic; Classification: Classification Problem, Classification Models; Trees: Tree models, Decision trees, Classification trees, Regression trees, Random forests; Time-Series Prediction: Time series modelling and prediction; Validation Techniques: Model validation, Out of sample validation, Cross-validation methods; Representation Learning and Goodness Measures: Dimensionality Reduction, Decorrelation, Clustering, Bias vs Variance trade-off, Uses of the Confusion matrix

15 Credits

This course focuses on how to build mathematical and statistical models that capture the essentials of real-world data science problems. The main aim is to provide students with modelling skills to practical problems with an emphasis on different modelling choices and their effects on solvability and planning under uncertainty. Topics include:

Linear programming: The structure, assumption, formulation, and the solution of linear programming problems with suitable examples Graphical interpretation of a linear programming problem with two decision variables Displaying and solving linear programming models on a Spreadsheet; Solving linear programming problems using the simplex method: The essence of the simplex method, setting up the simplex method, algebra of the simplex method, and simplex method in tabular form Spreadsheet Solution for Linear Programming Problem; Duality theory, and sensitivity analysis: The essence of duality theory, economic interpretation of duality, primal-dual relationships, examples and interpretations. The essence of sensitivity analysis, the role of duality theory in sensitivity analysis, applying sensitivity analysis to linear optimization problems; Integer Programming: Models with integer programming formulations, Gomory’s Cutting Plane Algorithm, A Branch and Bound Algorithm. Spreadsheet solution for integer programming problems; Transportation and Assignment problems: Characteristics of transportation problems, modelling variants of transportation problems, applications. Characteristics of assignment problems, modelling variants of assignment problems, applications; Network optimization problem: The terminology of networks, the shortest path problem, the minimum spanning tree problem, the maximum flow problem, the minimum cost flow problem, the network simplex method

5 Credits

With the increase in the use of technology, the amount of unstructured data available through public means have increased significantly. Hence, mining these unstructured public sources of data and obtaining information in an intelligence context has now become a modern-day requirement. This course introduces the aspects of OSINT from a data science perspective. This course focuses on the following topics:

Understanding Open Source Intelligence (OSINT) objectives and techniques as an intelligence-gathering discipline; Operational security, staying anonymous on the Internet; Software tools and online services usable for gathering, processing, and analysis of raw data; Evaluating and classifying data retrieved through OSINT gathering techniques.; Image and photo tracking, tracing, and analysis. ; Locating and verifying social media profiles and identities; Acquiring data from protected social media accounts. ; Cryptocurrency transactions and blockchain analysis; The Dark Web.; Counterintelligence techniques used by the criminal elements.

3 Credits

The dominance of textual information on the Internet and other digital sources has made mining high-quality information from text a critical task in modern-day data mining. The knowledge obtained via mining and analysis of textual data makes an impact on many areas such as business intelligence, social behaviour analysis and sentiment identification. In this course, the core focus will be as follows:

Introduction to Text Analytics, Treating Text as Data – Feature; Nature of Text, Document representation, Culture and Concordance; Text & Interaction, Content Analysis, Frequency Analysis, Co-Occurrence and N-grams; Information extraction: term extraction, named entity recognition, relation extraction, fact and event extraction; partial analysis vs. full analysis; Natural Language Processing, Part of Speech Tagging, Information Extraction and Dependency Parsing; Ethics in Text Mining, Resources for text mining: annotated corpora, computational lexica, ontologies, computational grammars; design, construction and use issues; Classification, Clustering: K-Nearest Neighbors, Simple Predictive; Topic Modelling and Dimensionality Reduction, Issues in large scale processing of text: distributed text mining, scalable text mining systems; Sources of text and social network data, Techniques to pre-processing text and social network data, The key metrics used in social network analytics (e.g. degree centrality, betweenness centrality, closeness centrality, etc.), Classification with social network techniques; Applying text mining and social network techniques in social media analytics

5 Credits

The core intention of the individual project is for the student to identify a contemporary research problem in data science or business analytics and to advance the problem via the scientific method under the guidance of a subject expert.

2 Credits

The objective of this course is to discuss the aspects of big data and its evolvement in the modern-day context. The core focus will be on each aspect of big data, their implications on the modelling techniques, algorithms and big data, identifying modern-day trends and technologies used to manipulate and store big data. This course outline is as follows:

Fundamentals: Differentiation between different big data models (5v’s, 7 v’s), Difference between common data analytics tasks and big data analytics tasks, challenges in modern-day data, streaming data; Software use for Analysis, Assessing the Quality of Big Data Analysis, Application Areas: Image Analysis, Recommender System, Speech Recognition Systems, Streaming Data

2 Credits

Processes are an essential part of any organization that strives continuously improve them. The improvement of the processes is usually motivated by inadequacies of the process or to obtain financial benefits. This course focuses on process improvement originating from data analytic models and reports. The main topics include:

Introduction to Analytics for Process Improvement: Process mining, types of process mining, How does process mining relate to data mining and analytics, Types of analysis methods; Process Models and Process Discovery: Using event data in process management, Event logs and process models; Different Types of Process Models: Four quality criteria for process discovery, Representation bias of process discovery, Business process model and notation, Process mining in the large; Process Discovery Techniques and Conformance Checking: Two-Phase Process Discovery And Its Limitations, Alternative Process Discovery Techniques, Conformance Checking: Positive and Negative Deviants, Conformance Checking, Exploring event data; Enrichment of Process Models: Mining Decision Points, Discovering Data Aware Petri Nets, Holistic Process Mining: Integrating Different Perspectives, Mining Bottlenecks, Mining Social Networks, Comparative Process Mining Using Process Cubes, Refined Process Mining Framework; Operational Support and Conclusion: Operational Support: Detect, Predict and Recommend, Guidelines for Logging, Process Mining Software, Mining Lasagna Processes, Mining Spaghetti Processes, Process Models as Maps, Data Science in Action.

2 Credits

Intelligent agents are simple processes operating in multifarious environments and circumstances. With the availability of large repositories of data belonging to gaming such as gambling, video games there is a significant incentive to develop intelligent agents whose actions, decisions and capabilities-driven by data. The outline of the course includes:

Agent Modeling: Fundamentals of player modelling, Opponent modelling, Distinguishing player modelling and opponent modelling; Types of Agent Models: Data-driven agent models, Action-based models, Learning techniques for Agent Models; Gaming and Agents: How to design intelligent agents, Single-agent systems vs multi-agent systems.

5 Credits

The independent study enables the students to put into practice the concepts studied throughout the entire course to solve real-world problems. This course consists of a number of real-world problems given to students as tasks in which they will use the knowledge gained through the course to solve these real-world problems.