February 18, 2016
Machine Learning Brings Infinite Options to Business Analytics
Leveraging Machine Learning brings untold opportunities to businesses seeking to garner the most value out of the massive amount of data they generate today.
Machine Learning (ML) is bringing forth countless opportunities for businesses looking to automate many IT processes, while also gathering insight into the massive amounts of data flooding the typical organization. While that may sound a bit like an affirmation of the capabilities of an intelligent machine, the fact of the matter is that ML is little more than complex algorithms that enable a system to learn from processing data. ML excels in several areas, especially those where a system can observe data transactions and then identify patterns based upon activity.
Ultimately, the machine learns how to correlate related activities and score those associations based upon intelligent algorithms. A process that proves useful for ideologies such as reputation scoring,
Dataiku – the software developer behind Data Science Studio (DSS) – is disrupting the predictive analytics market with an all-in-one predictive analytics development platform that gives data professionals the power to build and run highly specific services that transform raw data into business impacting predictions.
More than 50 customers in industries ranging from e-commerce, to industrial factories, to finance, to insurance, and pharmaceuticals use DSS on a daily basis to collaboratively build predictive dataflows to detect fraud, reduce churn, optimize internal logistics, predict future maintenance issues, and more. Dataiku has offices in Paris and New York.
Dataiku raised $3.7 million last year from two investors to grow its sales and tech team and international development initiatives.
predictions, targeting, recommendation engines, and process optimizations. When applied to business processes, such as IT security, ML can effectively gauge whether or not a particular activity is fraudulent in nature, or fits an acceptable usage pattern. Other advantages offered by ML include the burgeoning data analytics market, where automation powered by ML can uncover trends and relationships in data sets far too complex or large for a human to deal with. What’s more, ML accuracy increases with use, meaning the more an adaptive algorithm powered by an intelligent machine is used, the better it gets at doing its job.
Incorporating Machine Learning into business applications
Yet, a major challenge remains when it comes to implementing ML into analytics platforms, a challenge that is bracketed by complexity and integration. With those two detrimental factors in mind, analytics software vendor Dataiku has focused on a practical way to implement ML into analytical applications by seamless integrating the open source machine learning frameworks (H2O, Scikit-Learn, R, etc…) into its fully end-to-end advanced analytics software platform.
In the past, incorporating ML into any type of predictive analytics project required extensive coding and the creation of complex algorithms by data scientists intimately familiar with the nuances of statistics and mathematical principles. This situation limited the benefits offered by ML to businesses with very deep pockets and massive amounts of processing power, giving those businesses what some may deem as an unfair advantage, at least when it came to garnering insight into business trends. However, with ML now within the reach of most any business, those unfair advantages will likely evaporate and bring predictive analytics onto a level playing field.
“Incorporating machine learning algorithms into Data Science Studio allows us to provide an even more powerful all-in-one advanced data analytics package that our users can leverage to design, build, and deploy predictive applications, quickly. The DSS interface allows high performance machine learning technology to be easily utilized by beginner to expert data analysts,” said Florian Douetteau, Dataiku’s CEO.
Parkeon uses DSS and Machine Learning technology to revolutionize urban parking
Parkeon, a global leader in urban mobility, relies upon Data Science Studio to revolutionize urban parking with Path-to-Park, the first mobile application that predicts available parking spaces in real-time. The Path to Park project is based on the processing of massive volumes of heterogeneous data to forecast parking pressures in real-time with machine learning algorithms.
“Thanks to the seamless integration of H2O and scikit-learn in DSS, we’ve been able to deploy a workflow that goes directly from our production parking meter data to an api for the mobile application that applies distributed machine learning continuously on a growing volume of data,” said Yves-Marie Pondaven, CTO Parkeon.
Dataiku brings the power of Machine Learning technology to data analysts and data scientists of all skill levels
By integrating Machine Learning technology into DSS, Dataiku keeps true to the philosophy of bringing the best Big Data technologies together and making them available to all types of users – from beginner data analysts to expert data scientists or engineers – in one interface.
With DSS, Machine Learning becomes part of the whole flow – from raw data to predicted data. Finally, building and deploying powerful, predictive services from design to production is no longer a painful and overly time-consuming process.