Big Data Analytics
The ability to store, aggregate, and combine huge amounts of data, and perform insightful analytics on the results, has finally become more accessible and cost-effective.
Xanadu’s Big Data framework is built around an understanding of business mechanics, analysis of the business strategy and identifying value in unstructured and structured data. We specialise in high performance, cost-effective data mining, predictive analysis and data processing using techniques from traditional AI and machine learning.
- Identify the current Business Model and its KPI’s.
- Identify the Business Mechanics and the data that drives and represents the process.
- Define and identify the 4 V’s – Volume, Variety, Velocity and Value in the context of the available data set.
- Formalise a data management plan and data processing plan.
- Refine Business Strategy based on the data analysis.
Data Tools Used
- MapReduce/Big Query – Data processing framework
- R, Python – Data Mining, Predictive Analysis
- QlikView, Tableau, Pentaho – BI, Visualisation
Data Processing & Output Methods
- Parallel computing, GPU processing
- Date Warehousing – ETL, BI, Dashboards and KPI measurement
Data Analytics Methods
- Data Mining - Regression, clustering and classification. We also leverage dimension reduction, collaborative filtering, association rules and standard information retrieval methods.
- Predictive Analysis - Non-parametric regression and classification, Bayesian methods, time-series analysis, random forests, ensemble methods.
- Decision Analysis - Decision trees, sensitivity analysis and simulation.
- Optimisation - Integer and mixed-integer programming, stochastic and deterministic.