A high performance solution for predictive analytics

Informations projet

Neural Designer

N° de convention de subvention: 672676

DOI

10.3030/672676

Projet clôturé

Date de signature de la CE 13 Avril 2015

Date de début 1 Mai 2015

Date de fin 31 Octobre 2015

Financé au titre de

• INDUSTRIAL LEADERSHIP - Leadership in enabling and industrial technologies - Information and Communication Technologies (ICT)

Coût total

€ 71 429,00

Contribution de l’UE

€ 50 000,00

Coordonné par

ARTIFICIAL INTELLIGENCE TECHNIQUES SL
Spain

Periodic Reporting for period 1 - Neural Designer (A high performance solution for predictive analytics)

Période du rapport: 2015-05-01 au 2015-10-31

Résumé du contexte et des objectifs généraux du projet

In today’s digital world, the volume, variety and velocity of information generated by organizations are increasing significantly. A tool capable of performing intelligent analysis of that data will allow companies to make more effective decisions.

Artelnics develops and commercializes the deep learning solution for predictive analytics called Neural Designer. Deep learning is the most powerful method to discover intricate relationships, recognize complex patterns and predict current trends from data.

Some of the areas in which Neural Designer has been successfully applied are engineering (performance optimization, predictive maintenance ...) marketing (customer segmentation, product propensity ...) and health (medical diagnosis, microarray analysis ...).

The major challenge for machine learning is to achieve a solution that is capable of deducing the underlying causes and selecting the most relevant variables for a given data set. Therefore, the main ambition of Neural Designer is to include an innovative model selection framework capable of representing very complex data sets.

But for model selection algorithms to be efficient, Neural Designer needs to support big data and big computing technologies. It is therefore necessary to implement connection to data servers and parallelization in super-computing clusters, respectively.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

Within SME Instrument Phase 1, we have taken a step forward in the technical, commercial, legal and financial aspects of Neural Designer.

In the technical part, we have added new members to the development team. A new version, more optimized and plenty of innovations, has been released. Also, we have already developed different model selection algorithms to automatically find out which are the most relevant variables in a data set.

In the commercial part, an extensive analysis of the predictive analytics market has been carried out and the strategies needed to achieve our business goals have been defined. We have increased the sales team and are preparing a strong marketing campaign to advertise Neural Designer in our key target markets.

In the legal part, we have developed an effective strategy to manage and protect the intangible assets of Artelnics. This plan sets out a blueprint for a number of initiatives that the company will take in various areas, including intellectual property rights, trademarks and confidentiality.

In the financial part, sources of funding, sales forecast, expenses budget, cash-flow statement, return on investment and sensitivity have been analyzed in detail. The budget required for the successful development of Neural Designer has been calculated, and the feasibility of the project confirmed.

This study will allow Artelnics to apply for Phase 2 support, in which Neural Designer will be developed to become the most disruptive tool on the predictive analytics market.

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

Many applications of predictive analytics, from customer segmentation to medical diagnosis, arise from complex interactions that occur between different types of variables. A software tool which untangles these factors efficiently would be a great novelty likely to disrupt the existing predictive analytics market.

Neural Designer shall overcome all that difficulties. It will be able to guarantee the best possible results, by offering innovative model selection techniques and high computational performance. The immediate benefit for our customers is that our product will strength their businesses, allowing them to be more competitive.

Some potential impacts from the application of predictive analytics in engineering are reducing gas emmissions or miminizing equipment faults. In business, machine learning aligns products with customers and improves buying experience. In health, data mining can help to develop better diagnosis and more effective treatments.

Neural Designer will combine the best techniques from artificial intelligence and computer engineering to build the most powerful predictive models from large-scale data sets, thereby allowing the development of awesome predictive analytics applications.

neural-designer-start-page.png

Neural Designer logo