95830
We are sorry, this event is in the past, registrations are now closed. To view forthcoming events please see the Next Sessions tab or see our Schedule.
Big Data and Predictive Analytics are seeing explosive growth. There is a shortfall of individuals who understand these concepts. Not having these skills in any industry which aggregates data leads to an enormous competitive disadvantage.
This course will develop your skills in data analysis, summary statistics, forecasting and predictive analytics as well as predictive analytics using regression, probability, decision trees and neural networks. R has become the standard for statistical analysis and predictive analytics. The cost efficiency of the software – it is free – coupled with the incredible variety of supplementary packages available to augment its powerful script based language, makes it a one stop shop for all predictive analytics needs. It is little wonder that it is unseating SAS as the corporate choice for new Predictive Analytics projects. When taken together with the wide array of packages available to R, nearly anything required of an analyst practicing predictive analytics, is available in R and therefore, without any license costs. There are however three, entirely elective, GUI licensed products which provide such immeasurable value, beyond what is available in R, such that they are also covered in the course, this is Norsys Netica, Exhaustive and Neurosolutions Infinity. The course aims to bring Predictive Analytics together using R primarily, as well as some superb GUI licenced tools, to create a truly enterprise solution for all predictive analytics needs.

Predictive Analytics - St Julians

About Predictive Analytics with GUI Tools

Big Data, Predictive Analytics and Cognitive Analytics are seeing explosive growth. There is a shortfall of individuals who understand these concepts. Not having these skills in any industry which aggregates data leads to an enormous competitive disadvantage.

This course will develop your skills in data analysis, summary statistics, forecasting and predictive analytics as well as predictive analytics using regression, probability and neural networks.

About the Trainer

Your Trainer: RC is responsible for developing a proprietary predictive analytics platform that is used in a variety of industries.
The platform is predominantly positioned in Funds (Banks, Hedge Funds, Sovereign Wealth Funds and Mutual Funds etc.) to help manage investment portfolios using a blend of Quantitative Methods (Regression and Neural Networks typically) and Qualitative Methods (Expert Judgement and increasingly Augmented Bayesian Networks).
RC has over 16 years’ experience working with predictive analytics across a variety of industries and so can provide meaningful insight into the application of all techniques taught on the course in other problem industrial domains (e.g. Marketing Churn Models).

content test

content test

content test

content test

None

Testimonials

Hundreds of delegates from all over the world have attended LiquidNexxus training sessions. Here are some testimonials related to this course/event.

View More Testimonials
None

Agenda

Predictive Analytics Introduction

An introduction to big data and predictive analytics to provide foundation for the techniques to be covered on the course

Case Study: Elicitation of Variables

Create an exhaustive list of reasons that a customer may leave a telecommunications provider for the purposes of these variables being modelled by the company’s data analysts, then be carried through to predictive analytics.

Basic Statistics with Palisade StatTools

Predictive Analytics is rooted almost entirely in statistics and an understanding of basic statistics is vital to fully understand the linear and nonlinear predictive analytics techniques to be taught in the course

Abstraction and Transformations

Statistics underpin predictive analytics, however, the real power is the shaping and moulding of data for a given problem domain to create datasets that are more enriched and meaningful and thus lead to better predictive analytics models, measured by better predictive or classification accuracy

Logistic Regression

Logistic Regression is intended to model yes \ no, rather binary, type problems and in this regard, is useful for classification and behavioural analytics

Case Study: Logistic Regression

Using a dataset of fraudulent transactions develop a logistic regression model where the best classification accuracy will triumph.

Case Study: Bayesian Network

Create a Bayesian Network which has a better classification accuracy than that observed using Logistic Regression. Seek improvement in the model by transitioning through a Naïve Bayesian Network through to a Hierarchical Bayesian Network, before lastly looking at how variables that may not directly relate to the likelihood of default, relate to one another which when taken together may improve the overall accuracy.

Norsys Netica and Bayesian Analysis.

Bayesian Networks are an extremely powerful, yet highly intuitive and explainable, predictive analytics technique. Bayesian Networks not only allow prediction of the likelihood of an event happening but also provide a means of explaining the most probable environment that caused an event to happen

Neural Networks

Neural Networks are an extremely accurate, albeit complex and internally unexplainable, means of creating predictive analytics models. Neural Networks work just as well for classification as they do numeric prediction

Venue

None

Next Sessions

No session planned yet for this course, please click here to contact us.
Inhouse and eLearning available and we are always interested in working with new partners.