MSc Data Science

This course provides a broad introduction to machine learning and statistical pattern recognition. The very general topics will include supervised learning (generative/discriminative learning, parametric/non-parametric learning), unsupervised learning (clustering, dimensionality reduction), and learning theory (bias/variance tradeoffs).

The course will also discuss recent applications of machine learning (e.g., in computer vision, or other applications relevant to the research orientation of the department of computing).

Tutor: Dr Jamie Ward

An in-depth study of scalable solutions to manage, process and analyse Big Data on servers, clusters of computers or on the cloud. In particular you will study Big Data computing approaches, trends and technologies as Apache Hadoop based on the MapReduce scalable computing approach, HBase database system and NoSQL technologies, Hive data warehouse system, Pig Latin for productively creating large scale data applications, Mahout for scalable machine learning, and scripting languages as Python currently used within Big Data processing.

You will be exposed to and develop various types of Big Data applications including  social network mining, reality mining, mobile phone large data analysis, intelligent web, etc.

This module introduces programming for Data Science, concentrating primarily on the tools and techniques that are key to achieve results quickly. The module covers current programming languages and environments commonly used in the wider Data Science community, along with ancillary tools and software systems, and gives the student the foundational skills to allow them to develop data-related software for their specific areas of interest.

The knowledge and skills in this course cover the following general areas: Data representations: basic data types, comma-separated variables, Xtensible Markup Language, JavaScript Object Notation, Resource Description Framework, Relations; Data acquisition, storage, retrieval and publication: filesystems, version control, network programming, HTTP, Web servers, relational database systems; Data programming: string processing, numeric vector processing, data frames, scripting and statistical programming; Visualizations: automatically generating charts, graphs, and choropleth maps.

Tutor: Dr. Aaron Gerow

The module introduces students to research topics related to Data Science and Computational Social Sciences, and to modern applications of Machine Learning, Data Mining, Computational Statistics and Data Analysis, Statistical Computing, Forecasting, Big Data Management and Analytics, and Soft Computing in industry. 

In particular students are given talks by Data Science and Computing industry professionals and by academics whose monodisciplinary or interdisciplinary research in areas as Computer Science, Sociology, Psychology, Bioinformatics, Biomedical Statistics and other disciplines is based or involves Data Analysis. The module will guide students' work in exploring a research theme of interest, as a preliminary phase preparing them for the final project. 

The module will expose students also to learning techniques of conducting research work and of writing scientific reports. Moreover students will be taught approaches for devising and optimising their strategies adapted to their skills in seeking and securing an employment, whether in research or the industry sector. The module will also expose students to a discussion of the ethical and legal issues related to the area of Data Science and its applications.

Students undertake substantial independent research projects that will allow them to demonstrate project planning and management skills, research skills, and written and oral presentation skills. You will integrate your knowledge and skills acquired in previous modules to implement a final project in Data Science or related interdisciplinary topics. The work will consist of a combination of research and highly applicative elements in various proportions.

Students will have the opportunity to join research groups at Goldsmiths whose work is within the Data Science field, or is interdisciplinary and makes use of methodologies from Data Analysis or other components of Data Science as instruments of research. Project work can also be developed in industry as an internship.

This module aims to cover the key statistical concepts and techniques for data analysis. The areas covered include descriptive statistics, probability theory, Bayes theorem, naïve Bayesian classifier, regression, prediction, and classification. It covers further topics in statistical data mining like K-Nearest Neighbors and Tree Models for Regression and Classification, Bagging, Random Forest and Boosting Techniques.

A machine is artificially intelligent when it manages to perform a task that we thought, until the machine proved capable, required human intelligence. Afterwards, we recalibrate our definition of intelligence.

Ai is a broad field and includes many disciplines and ideas. But a new technique is taking over.
Machine learning, an AI technique, has been around for a while. A special machine learning practice known as Deep Learning is revolutionising AI. It is everywhere - or will be soon.

Simply, AI ≈ DL.

We will learn how to build DL programs - known as models - and train them on huge datasets. We will be using TensorFlow, Google's important DL resource. TensorFlow, in turn, is programmed using Keras, a high-level Python library. We will write Keras DL code in Jupiter notebooks and plot graphs with another Python library, matplotlib. Finally, our programs will rely on Python's special library for numerical calculation - Numpy

This module provides you with advanced skills in designing interactive systems and an in-depth understanding of emerging practico-theoretical developments in interaction design.

The module is delivered as a series of workshops, lectures and seminars where you're introduced to a range of key technical skills for making interactive platforms, and develop an understanding of the role of prototyping though the embedding of technical work in the pursuit of a series of design briefs.

You'll be able to then use these technologies in your projects, and develop an understanding of the roles of software and hardware development.

Introduces the theory and practice of neural computation. Covers the principles of neurocomputing with artificial neural networks widely used for addressing real-world problems such as classification, regression, pattern recognition, data mining, time-series prediction. We look at supervised and unsupervised learning. We study supervised learning using linear perceptrons, and non-linear models such as probabilistic neural networks, multilayer perceptrons, and radial-basis function networks. Unsupervised learning is studied using Kohonen networks. We provide contemporary training techniques for all these neural networks, and knowledge and tools for the specification, design, and practical implementation of neural networks.

Tutor: Dr Nikolay Nikolaev

This module aims to cover a wide range of applied data analysis techniques using R. With its strong focus on practical application of the techniques, the topics covered include data sources, data gathering, interpretation of statistical results, visualization, principal component analysis, practical implementation using R and decision making based on data analysis.

A large amount of data is available in electronic resources, both offline and online. This module will give a broad introduction to techniques for gathering data from electronic sources, such as databases and the internet. It will cover both fundamental ideas and the use of some of the most important currently available tools. The module will also present tools and ideas for more effectively using the internet to communicate, visualise and generate news stories. 

This module focuses on the geometry and statistical processing underpinning component analysis methods, including numerical and score data and other quantitative data, as well as qualitative measurements and textual data. Classification, (hierarchical clustering, partitioning and discriminant analysis) are also relevant. The module also looks at the interpretation of results, graphical displays and other outputs, and practical implementation using the open source and universally supported R statistical and visualisation environment.