In this section
The Centre for Intelligent Data Analytics (TCIDA) deploys AI to business and investment issues.
There are many state-of-art computer labs around the world already deploying AI on very complex academic problems; for example, strategic games (Google/DeepMind’s AlphaZero (Go, Chess, Shogi)), virtual assistants (Apple's Siri), autonomous vehicles (Tesla Autopilot, Google Waymo, FiveAI), and natural language processing (IBM’s Watson).
Significant progress in AI has been made through these advancements over the last few years. However, the systems highlighted above, which, with good cause, have attracted the media spotlight, are built for well defined homogeneous and discrete environments, where the data at hand has a specific and unambiguous meaning. Commonly, business data environments are a far cry from these, having heterogeneous systems accessing siloed data stores with many textual fields that may be corrupt, misplaced, and sometimes missing. Commercial and legal constraints require ever greater needs for process transparency, data protection, and accountability. Businesses want high levels of process automation and interoperation at affordable costs and minimal risk.
What is AI?
In 1955 John McCarthy first coined the term Artificial Intelligence to mean:
‘.. the science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable.’
On the simplest level, one may read “decision making” for “intelligent”. Rising up the scale one may consider decision quality (optimality), goal orientation (strategic), context dependence (adaptability), decision effect (impact), and semantic grounding (meaning).
At TCIDA we use the term Analytics to imply a multi-disciplinary approach to:
The discovery, interpretation, and communication of meaningful patterns in data.
This typically involves the interrogation of [often large] databases, interpretation against ontologies, and using tools derived from artificial intelligence, machine learning, computer science, maths, statistics and operational research.
In addition, the application of these technologies to business concerns (especially, automated decision making) raise numerous legislative and ethical issues; including bias, accountability, the opaqueness of AI processes, data privacy, collusion, and unintended consequences affecting the market.
"The TCIDA's mission is to be the internationally preeminent centre for the application of Artificial Intelligence (AI) and Analytics to business processes and business problems."
The tools TCIDA uses
TCIDA has world-class expertise in AI including a proven track record in using the following technologies in the design of complex Intelligent Systems:
Machine Learning: Artificial Neural Networks (ANNs); Deep Learning (DL); Reservoir Computing (RC); Support Vector Machines (SVMs); Graphical [Bayesian] Models (GM): Hidden Markov Models (HMMs), Conditional Random Fields (CRFs), Partially Observable Markov Decision Processes (POMDPs) and Bayesian Networks (BNs); Swarm Intelligence (SI): Evolutionary MethodsAgent-based Automated Decision Processes; Particle Swarms Optimisation (PSOs) and Stochastic Diffusion Search (SDS).
Data Analysis: Classification and Clustering; Kernel Methods (KM): Radial Basis Functions (RBFs), Principal Component Analysis (PCA); Fuzzy Logic; Bio-informatics; Sequential and Markov Monte Carlo Methods; Hierarchical Time Series; Time Series Regression and Correlation.
Data Representation: Databases; Big Data Architectures; Knowledge Representation; Web-Spidering and Wrapper Extraction; Formal Grammars; Stochastic Finite State Automata; Semantic Methods; Natural Language Processing (NLP); Topic Modelling; Computational Linguistics; Sentiment Analysis; Vector and Tensor Semantics; Quantum Linguistics.
AI, Analytics and Thematic Investing
A recent Bank of America/Merrill Lynch report1 claims the Robot and AI market will grow to $153 billion with AI-based Analytics expected to explode into a $70 billion industry by 2020.
‘We estimate the robots and AI solutions market will grow to US$153bn by 2020, comprising US$83bn for robot and robotics, and US$70bn for AI-based analytics. We anticipate US$14-33tn in annual creative disruption impact in 10Y, including US$8-9tn of cost reductions across manufacturing and healthcare, US$9tn cuts in employment costs via AI-enabled automation of knowledge work and US$1.9tn in efficiency gains via autonomous cars and drones. Adoption of robots and AI could boost productivity by 30% in many industries while cutting manufacturing labour costs by 18-33%’ (Thematic Investing, 2015).
The report suggests that the market for AI-based Analytics could grow to US$70bn by 2020:
‘AI is already used in a wide range of applications including web search systems, marketing recommendation functions, the security business and financial trading programs. In the future, we expect the applications for AI to expand to include driverless cars and service robots. Research firm IDC forecasts the Internet of Things (IoT) market related to analytics, with AI technologies at the core, will grow by CAGR of 36% – from US$8.2bn to US$70bn – in 2013–20.’
The business problems TCIDA addresses
Procurement Analytics: Product Matching; Price Variance; Invoice Matching; Contract Adherence; Vendor Management; Product Consolidation; Classification; Spend Forecasting.
HR Analytics: Communication Network Analysis; Company Structure; Process Mapping; Actor Recipient Topic Models; Employer Wellbeing; Insider Threat Detection.
Process Automation: Robotic Process Automation; Extract, Load, Transform (ELT) Process; Web Scraping; Process Robustness; Know Your Customer (KYC) and Anti-Money Laundering (AML) Automation.
Data Quality and Provenance: Data Cleaning; Entity Recognition; Data Normalisation; Parsing; Knowledge Management; Data Enrichment; Schema Mapping; Data Warehousing.
Automation and AI Ethics: Data Governance; Data Bias; Profiling; Algorithmic Opacity; Process Audit; Accountability; Data Protection.