Predictive Analytics
Let us help you with predictive analytics for you to make predictions about unknown future events
Artificial Intelligence
It is capable of making abstract associations, finding patterns in unstructured data and interacting with the environment in a self learning manner. In recent years, artificial intelligence has been propelled by advances in algorithms development, increase in computing power and in data storage capacities.
Due to the complexity involved, exploring the benefits of integrating artificial intelligence remains aspirational for most companies, despite increasing investments.
Use cases include autonomous transport vehicles, immersive gaming experience and workforce automation.
Machine Learning
Is capable of describing complex data patterns, making prediction and recommending optimal action based on predicted outcomes. Two types of machine learning algorithms exists based on how they classify information: supervised learning find patterns which explain a clearly defined outcome (e.g. customer churn); unsupervised which find patterns in the data without a specific outcome (customer segments).
Machine learning algorithms are relatively easy to apply on most common types of data (e.g. retail transaction data) and thus are increasingly used to support business decision-making.
Use cases include customer behavioral segmentation, predicting probability of customer churn, forecasting demand and signaling fraudulent behaviour.
Deep Learning
A specialized branch of machine learning with specific set of algorithms designed to find patterns in highly unstructured data (e.g. video feed, written and spoken language). It is based on the concept of artificial neural networks which has two types: convolutional (CNN) and recurrent (RNN).
Convolutional Neural Networks are used to find patterns which make up an unstructured inputs (e.g. a letter in an image) and learn to detect its unique features, assuming that all inputs are independent from one another. Recurrent Neural Networks relax the assumption of input independence and incorporate the sequence of previous inputs as predictors of the value of the current one (particularly powerful in text and voice recognition).
Use cases include real time language translation, social sentiment analysis and interactive chatbots.
Big Data
It is best used as a collective term to describe specialized technologies and know how necessary to structure and extract meaning out of datasets which are many orders of magnitude larger than what the typical business user is accustomed to. The definition of Big data therefore a moving target as organizations get more sophisticated in handling their data management systems.
Besides it’s high volume, big data has the the property of high variety (multiple sources both structured and unstructured) and high velocity (frequency of inputs). Special technologies and skills are needed to manage such data systems,
Use cases for big data in include real time data tracking such as transactions at retail chains, airline ticket prices and stock market figures.
