Alex Guazzelli

Alex Guazzelli

Location
Greater San Diego Area
Industry
Information Technology and Services

As a LinkedIn member, you'll join 300 million other professionals who are sharing connections, ideas, and opportunities.

  • See who you and Alex Guazzelli know in common
  • Get introduced to Alex Guazzelli
  • Contact Alex Guazzelli directly

View Alex's full profile

Alex Guazzelli's Overview

Connections

500+ connections

Alex Guazzelli's Patents

Alex Guazzelli's Skills & Expertise

  1. Fraud Detection
  2. Predictive Analytics
  3. Predictive Modeling
  4. Data Mining
  5. Machine Learning
  6. Risk Management
  7. Cloud Computing
  8. Statistics
  9. Neural Networks
  10. Business Intelligence
  11. Credit Scoring
  12. Data Analysis
  13. Time Series Analysis
  14. Customer Analysis
  15. Data Visualization
  16. Big Data
  17. Statistical Modeling
  18. Business Analytics
  19. Analytics
  20. Artificial Intelligence
  21. Product Management
  22. Logistic Regression
  23. Enterprise Software
  24. Quantitative Analytics
  25. Segmentation
  26. Marketing Analytics
  27. Text Analytics
  28. Cluster Analysis
  29. R
  30. Optimization
  31. Decision Trees
  32. Agile Methodologies
  33. SAS programming
  34. Text Mining
  35. SAS
  36. Teradata
  37. Hadoop
  38. Data Integration
  39. Data Modeling
  40. Linear Regression
  41. PMML
  42. Data Management
  43. Data Warehousing
  44. Operations Research

View All (44) Skills View Fewer Skills

Alex Guazzelli's Languages

  • English

    (Full professional proficiency)
  • Portuguese

    (Native or bilingual proficiency)
  • French

    (Professional working proficiency)
  • Spanish

    (Professional working proficiency)

Alex Guazzelli's Publications

  • Representing predictive solutions in PMML: Move raw data to predictions

    • IBM developerWorks
    • September 28, 2010
    Authors: Alex Guazzelli

    PMML, the Predictive Model Markup Language, is the de facto standard used to represent a myriad of predictive modeling techniques, such as Association Rules, Cluster Models, Neural Networks, and Decision Trees. These techniques empower companies around the globe to extract hidden patterns from data and use them to forecast behavior. In this article, start with a look at the predictive modeling techniques that are directly supported by the standard. However, given that a predictive solution is more than the statistical techniques it harbors, then dive even deeper into the language and explore the transformations and functions that are used for data processing by illustrating the use of data pre-processing and modeling in PMML as it is used to represent a complete predictive solution.

  • What is PMML? Explore the power of predictive analytics and open standards

    • IBM developerWorks
    • July 30, 2010
    Authors: Alex Guazzelli

    The Predictive Model Markup Language (PMML) is the de facto standard language used to represent predictive analytic models. It allows for predictive solutions to be easily shared between PMML compliant applications. With predictive analytics, the Petroleum and Chemical industries create solutions to predict machinery break-down and ensure safety. PMML is supported by many of the top statistical tools. As a result, the process of putting a predictive analytics model to work is straightforward since you can build it in one tool and instantly deploy it in another. In a world in which sensors and data gathering are becoming more and more pervasive, predictive analytics and standards such as PMML make it possible for people to benefit from smart solutions that will truly revolutionize their lives.

  • PMML in Action: Unleashing the Power of Open Standards for Data Mining and Predictive Analytics

    • CreateSpace
    • May 18, 2010
  • Efficient Deployment of Predictive Analytics through Open Standards and Cloud Computing

    • ACM SIGKDD Explorations
    • July 2009
  • PMML: An Open Standard for Sharing Models

    • The R Journal
    • May 2009
  • HYCONES II: a tool to build Hybrid Connectionist Expert Systems

    • MD Computing, AMIA, Inc.
    • 1994

    ADAPA's ability to seamlessly integrate predictive analytics and business rules for the real-time execution of hybrid systems is remarkable.

    Back in academia in the early nineties, I was very fortunate to be part of a great team of researchers studying and building hybrid systems. HYCONES was a tool that combined neural networks and frames to solve classification problems.

  • Competitive Hebbian learning and the hippocampal place cell system: Modeling the interaction of visual and path integration cues

    • Hippocampus
    • June 2001
    Authors: Alex Guazzelli, Mihail Bota, Michael Arbib
  • Dissociation of the Effects of Bilateral Lesions of the Dorsal Hippocampus and Parietal Cortex on Path Integration in the Rat

    • Behavioral Neuroscience
    • December 2001
    Authors: Alex Guazzelli, Etienne Save, Bruno Poucet
  • Incorporating path integration capabilities in the TAM-WG model of rodent navigation

    • Neurocomputing
    • 1998
    Authors: Alex Guazzelli, Mihail Bota, Michael Arbib
  • Context-Dependent Reorganization of Spatial and Movement Representations by Simultaneously Recorded Hippocampal and Striatal Neurons During Performance of Allocentric and Egocentric Tasks

    • Behavioral Neuroscience
    • August 2004
    Authors: Alex Guazzelli, Oxana Yeshenko, Sheri Mizumori
  • Affordances, Motivations, and the World Graph Theory

    • Adaptive Behavior
    • January 1998
    Authors: Alex Guazzelli, Mihail Bota, Fernando Corbacho, Michael Arbib
  • A simplified ARTMAP architecture for real-time learning

    • Lecture Notes in Computer Science / Springer
    • 1993
    Authors: Alex Guazzelli, DANTE BARONE, Edson Carvalho Filho
  • NeWG: In search of the rat's World Graph

    • Journal of the Brazilian Computer Society
    • July 1997
    Authors: Alex Guazzelli, Michael Arbib
  • The PMML Path Towards True Interoperability in Data Mining

    • In Proceedings of the 17th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
    • August 2011

    As the de facto standard for data mining models, the Predictive Model Markup Language (PMML) provides tremendous benefits for business, IT, and the data mining industry in general, since it allows for predictive models to be easily moved between applications. Due to the cross-platform and vendor-independent nature of such an open-standard, auto-generated PMML code is often represented in different versions of PMML. A tool may export PMML 2.1 and another import PMML 4.0. This problem raises the issue of conversion. For true interoperability, PMML needs to be easily converted from one version to another.
    In this paper, we describe the capabilities associated with the “PMML Converter”. This application represents a great step in the PMML path towards true interoperability in data mining. Besides converting older versions of PMML to its latest, the PMML converter checks PMML files for syntax issues and, if issues are encountered, automatically corrects them.
    This paper also describes the capabilities associated with an interactive PMML-based application, the “Transformations Generator.” Auto-generated PMML code can omit important data pre-processing steps which are an integral part of a predictive solution. The Transformations Generator aims to bridge this gap by providing a graphical interface for the development and expression of data pre-processing steps in PMML.

  • Scorecard Element in PMML 4.1 Provides Rich, Accurate Exchange of Predictive Models for Improved Business Decisions

    • In Proceedings of the 17th ACM SIGKDD Conference on Knowledge Discovery and Data Mining
    • August 2011

    This paper illustrates the dedicated Scorecard element introduced in the 4.1 specification of the PMML standard, including the various design and computational options available for returning reason codes alongside each computed score. The paper is intended to help both producers and consumers of scorecards as PMML documents.

  • Predictive analytics in healthcare: The importance of open standards

    • IBM developerWorks
    • November 2011
    Authors: Alex Guazzelli

    As digital records and information become the norm in healthcare, it enables the building of predictive analytic solutions. These predictive models, when interspersed with the day to day operations of healthcare providers and insurance companies, have the potential to lower cost and improve the overall health of the population. As predictive models become more pervasive, the need for a standard, which can be used by all the parties involved in the modeling process: from model building to operational deployment, is paramount. The Predictive Model Markup Language (PMML), is such a standard. It allows for predictive solutions to be easily shared between applications and systems. This article describes the latest release of PMML, Version 4.1, and several ways it can be used to expedite the adoption and use of predictive solutions in the healthcare industry.

  • PMML in Action (2nd Edition):

    • CreateSpace
    • January 2012
    Authors: Alex Guazzelli

    The Predictive Model Markup Language (PMML) standard has reached a significant stage of maturity and has obtained broad industry support, allowing users to develop predictive solutions within one application and use another to execute them. Previously, this was very difficult, but with PMML, the exchange of predictive solutions between compliant applications is now straightforward.

    The aim of this book is to present PMML from a practical perspective. It contains a variety of code snippets so that concepts are made clear through the use of examples. Readers are assumed to have a basic knowledge of predictive analytics and its techniques and so the book is intended for data mining movers and shakers: anyone interested in moving predictive analytic solutions between applications, including students and scientists.

    PMML in Action is a great way to learn how to represent your predictive solutions through a mature and refined open standard. For the 2nd edition, the book has been completely revised for PMML 4.1, the latest version of PMML. It includes new chapters and an expanded description of how to represent multiple models in PMML, including model ensemble, segmentation, chaining, and composition. The book is divided into six parts, taking you in a PMML journey in which language elements and attributes are used to represent not only modeling techniques but also data pre- and post-processing.

    With PMML, users benefit from a single and concise standard to represent predictive models, thus avoiding the need for custom code and proprietary solutions.

  • Predicting the future, Part 1: What is predictive analytics?

    • IBM developerWorks
    • May 29, 2012
    Authors: Alex Guazzelli

    You can use predictive analytics to solve your most challenging problems. It helps you discover patterns in the past, which can signal what is ahead. This first article of a four part series focuses on predictive analytics. It starts by looking at analytics in general, and then positions data-driven analytics against business rules and expert knowledge. Both types of knowledge can enhance your decision-making ability. Predictive analytics is able to discover hidden patterns in data that the human expert may not see. It is in fact the result of mathematics applied to data. As such, it benefits from clever mathematical techniques as well as good data. Given that we can apply predictive analytics to a myriad of datasets in different industries and verticals, this article helps you identify a few applications of predictive analytics on your own.

  • Agile predictive analytics on IBM SmartCloud Enterprise

    • IBM developerWorks
    • June 7, 2012
    Authors: Alex Guazzelli

    Companies all over the world face a barrage of data (which is why it's referred to as "Big Data") that keeps getting larger and more complex — to extract value and insight from it, organizations are resorting to predictive analytics. By using statistical techniques that can discover important patterns present in past data, companies can tap into the future, allowing for more precise and consistent business decisions. But, to fully benefit from big data solutions, an infrastructure must be in place for agile deployment and execution. While an open standard such as Predictive Model Markup Language (PMML) allows for the instantaneous moving of solutions between development and operational environments (collectively known as DevOps), cloud computing and Software as a Service offer the power and flexibility necessary to bring them to life. In this article, the author introduces the concepts of agile predictive analytics on the cloud, provides some information on PMML, and offers a real-world example of a tool that integrates these elements under a single control.

  • Predicting the future, Part 2: Predictive modeling techniques

    • IBM developerWorks
    • June 19, 2012
    Authors: Alex Guazzelli

    This is the second article of a four part series focusing on the most important aspects of predictive analytics. Part 1 offered a general overview of predictive analytics. This article focuses on predictive modeling techniques, the mathematical algorithms that make up the core of predictive analytics.

  • Predicting the future, Part 3: Create a predictive solution

    • IBM developerWorks
    • July 3, 2012
    Authors: Alex Guazzelli

    This is the third article of a four-part series focusing on the most important aspects of predictive analytics. Part 1 offered a general overview of predictive analytics. Part 2 focused on predictive modeling techniques; the mathematical algorithms that make up the core of predictive analytics. This article describes how to use those techniques and create a predictive solution.

  • Predicting the future, Part 4: Put a predictive solution to work

    • IBM developerWorks
    • July 10, 2012
    Authors: Alex Guazzelli

    This is the last article of a four-part series focusing on the important aspects of predictive analytics. Part 1 offered a general overview of predictive analytics. Part 2 focused on predictive modeling techniques, the mathematical algorithms that make up the core of predictive analytics. Part 3 put those techniques to use and described the making of a predictive solution. This final article focuses on the deployment of predictive analytics, or the process of putting predictive solutions to work.

  • The R pmmlTransformations Package

    • ACM
    • 2013

    As the de facto standard for data mining models, the Predictive Model Markup Language (PMML) provides tremendous benefits for business, IT, and the data mining industry in general. Due to the cross-platform and vendor-independent nature of such an open-standard, it allows for predictive models to be easily moved between applications.

    Although PMML has offered support for common data transformations for quite some time now, the release of PMML 4.0 in 2009 brought the support for data pre-processing steps to a new level. As a consequence, several data mining tools and model building platforms have been adding more and more support for data pre-processing into the PMML code they export. It is no
    surprise then that the same is true for R.

    R has become a popular statistical platform for all things analytics. The R Project allows for a myriad of specialized packages to be installed and utilized by its users as needed. These include packages and functions for predictive analytics and model building. A package for exporting PMML out of several model types is also available. Called the pmml package, it allows for a few data pre-processing steps to be exported together with the modeling technique itself. However, a package to enable data transformations in a generic way was still missing.

    This paper describes a package which intends to close this gap. The pmmlTransformations package provides R users with functions that greatly enhance the available data mining capabilities and PMML support by allowing transformations to be performed on the data before it is used for modeling. The pmmlTransformations package works in tandem with the pmml package so that data pre-processing can be represented together with the model in the resulting PMML code.

  • Extending the Naive Bayes Model Element in PMML: Adding Support for Continuous Input Variables

    • ACM
    • 2013

    The Predictive Model Markup Language (PMML) is the de facto standard to represent data mining and predictive analytic models. With PMML, one can easily share a predictive solution among PMML-compliant applications and systems.

    PMML as a standard has evolved significantly over the years. PMML 4.1, the language’s latest version represents a major leap forward in terms of its ability to represent data post-processing and multiple models. It also provides entirely new model elements for supporting Scorecards and K-Nearest Neighbors. The same is no exception for PMML 4.2, currently being worked on by the Data Mining Group (DMG), the body responsible for maintaining and advancing the PMML standard. PMML 4.2 is bound to offer new elements and increased capabilities. This article describes one of such improvement. In particular, it proposes extending the existing model element for Naïve Bayes Classifiers to support continuous input fields.

    The R Project is a popular choice for data miners to analyze and build predictive models. Naïve Bayes is just one of a myriad of model types supported by R. The R e1071 package provides a
    naiveBayes function to build Naïve Bayes Models using categorical as well as continuous fields. The R pmml package has been recently extended to allow for the export of PMML code for objects built with the naiveBayes function. For now, it includes a PMML Extension element for continuous fields, but with the release of PMML 4.2, the support will be standardized. This article describes this process in view of our proposal to extend the current model element for Naïve Bayes Models.

  • Representing Model Ensembles in PMML

    • useR! 2014 Conference
    • July 1, 2014

    The R pmml package is now able to export PMML for ensemble models via the ada and randomForest functions. In this presentation, we describe all the steps necessary to export random forest and stochastic boosting models from R into PMML and show how the PMML standard is capable of representing not only model ensembles but also any R specified treatments for missing and invalid values as well as outliers. Additional functions available to the data scientist through the R pmml package include the ability to perform data pre- and post-processing.

  • The pmmlTransformations Package

    • useR! 2014 Conference
    • July 1, 2014

    The R pmmlTransformations package implements many of the commonly used transformation operators used by data scientists, among them the Z-transform, linear transformation, data discretization, data normalization and value mapping. The result is not only the transformed data itself but also information to represent the transformation operators in PMML format.

View Alex Guazzelli’s full profile to...

  • See who you and Alex Guazzelli know in common
  • Get introduced to Alex Guazzelli
  • Contact Alex Guazzelli directly

View Alex's full profile

Viewers of this profile also viewed...