text stringlengths 0 1.75k |
|---|
Grading Type: |
N |
Prerequisite Modules: |
Rationale and Purpose of the Module: |
Conversational Artificial Intelligence is the software and processes by which speech is transformed into input for computers and smart devices. |
This module will provide students with an overview of the programming foundations that are used to engineer modern Conversational AI (CAI) systems. |
Syllabus: |
An introduction to scripting Languages and Environments for Scientific Computing: An introduction to the syntax of one modern scripting languages (e.g. Python, Julia or the latest equivalent) and environments. |
An introduction to Numerical issues in CAI systems. The bag of words test and the generation of analysis vectors. Numerics support in typical scientific scripting (e.g., Numpy/Scipy). Matrices and linear algebra |
Graphics and Scientific Visualization of words and sentences in CAI systems: Using scripting languages to build scientific visualizations (scalar, vector fields). |
Random Numbers and Probability: Random number generation: |
Classification in CAI systems. Simple classifiers. K-means. Linear classifiers: Perceptron. Least squares and gradient descent. Modern optimization for neural networks: Nesterov momentum, the ADAM optimizer. |
Learning Outcomes: |
Cognitive (Knowledge, Understanding, Application, Analysis, Evaluation, Synthesis) |
On successful completion of this module, students will be able to: |
Given a target programming language, demonstrate proficiencies in the syntax necessary to implement standard programming constructs in CAI systems. |
Given a basic bag of words dataset, construct simple programmes to perform simple analysis operations. |
Given a CAI problem, identify and evaluate the outputs through appropriate visualisation. |
Given a CAI problem, discriminate and select basic approaches to scientific computing. |
Given an appropriate bag of words data set, the student will write a program to process the data e.g. find the principal components. |
Affective (Attitudes and Values) |
On successful completion of this module, students will be able to: |
Given datasets, demonstrate knowledge of how to question whether the data is representative and how to attempt to address any biases. |
Given a bag of words to investigate, demonstrate knowledge of how to identify and discuss any potential ethical considerations that might obtain. |
Through the use of appropriate outputs including visualisation, demonstrate ability to |
identify a basic intent from a bag of words dataset. |
Psychomotor (Physical Skills) |
N/A |
How the Module will be Taught and what will be the Learning Experiences of the Students: |
Materials will be delivered in a blended manner through weekly pre-recorded sessions and live class sessions. The module material will include video recordings as well as readings, exercises, and assignments. The focus is on the reduction of theory to practice so there will be a strong emphasis on developing participa... |
Research Findings Incorporated in to the Syllabus (If Relevant): |
Prime Texts: |
Langtangen (2016) A Primer on Scientific Programming with Python , Springer |
Other Texts: |
Beazley (2016) Machine Learning , Cambridge |
Goodfellow & Bengio (2014) Deep Learning , MIT Press |
Lane Hobson (Author), Howard Cole (Author), Hapke Hannes (Author) (2019) Natural Language Processing in Action: Understanding, analyzing, and generating text with Python , Manning |
Programmes |
Semester(s) Module is Offered: |
Spring |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.