New trends in Machine Learning
Organizations are late to adopt innovations in machine learning - in spite of dramatic changes in business strategies and practices.
In recent years, new abilities of storing, processing and analyzing huge amounts of diverse data, with lower costs and reduced convoluted demands on management, are shaking-up the classic data warehouse realm. Modern Visualization tools are taking over the complex, rigid and time-consuming legacy business intelligence processes. These new tools are so user-friendly they succeed in making the analysis process accessible to business level operators.
Furthermore, cloud computing removes another obstacle from organizations who want to implement business analytics tools without the concern of frequent hardware and software platform upgrades. Financial and organizational barriers to the world of business analytics are now lower than ever.
Nonetheless, it appears that this revolution skips over predictive analytics, and in this area organizations still rely on legacy processes. Predictive analytics practices in consumer-oriented organizations, used for planning per-customer marketing resources allocation, cross-sell/up-sell prediction, churn prevention, customer loyalty and lifetime value optimization have not significantly changed in the last decade. Aside perhaps from retitling job openings for Statisticians as Data Scientists…
The problem is that these new Data Scientists are using available data-mining, machine learning and trend discovery tools to build statistical models in the same old manual and time-consuming way as statisticians did ten years ago. Like then, they put weeks into manual design, development, updating and upgrading their models, trying to fit them to the rapidly changing business environment, new sources and new types of data. Implementing these models is also a complex process which often requires large hardware and software investments. These constraints create a gap between the dynamic marketing requirements of customer-oriented businesses and the statisticians and analysts who need to respond to these needs.
As a result we see that predictive analytics strategies and applications are not gaining the popularity they merit while corporations already using them need to keep paying the elevated costs of maintaining such manual modeling techniques.
New wave machine learning
G-STAT BRAINs™ [Big-data Recommendations for Actionable Insights] applications transform big data automatic machine learning into an accessible and affordable daily marketing tool. They come preloaded with 200 out-of-the-box ready to deploy data-management, machine learning and trend discovery processes. Multi-segment-modeling projects are easily configured to the organization's sales and marketing processes. Marketing users can now design and deploy thousands of sophisticated end-to-end models for churn prediction and prevention, next-best-action, cross-sell, up-sell, revenue impact prediction and lifetime value optimization in hours instead of months. Risk and data security analysts can now design and deploy sophisticated end-to-end models for credit scoring, collection scoring, fraud management, anomalies detection and money laundering scoring in hours instead of months.
Consumer oriented organizations such as banking, finance, retail and telecom can now reduce development of individually targeted marketing campaigns time-to-results from months to hours. They can provide their business, marketing and risk management teams with tools that deliver higher predictability results that immediately impact the bottom line. TCO is reduced across the board with lower infrastructure [software and hardware] and reduced labor [statisticians and analysts]. The G-STAT solution may be the innovation that machine learning market was waiting for, in order to regain its prime position in the world of customer-oriented marketing & risk and fraud management.