Showing posts with label Algorithms. Show all posts
Showing posts with label Algorithms. Show all posts

Monday, June 08, 2009

McKinsey on Risk

McKinsey Consulting are providing a 14 minute video of Peter L. Bernstein talking about risk. This is a fascinating and valuable video that I promise you will learn one or two things about risk. It's unfortunate that the video marks the passing of Mr. Bernstein, but I have put his book "Against the Gods: The Remarkable Story of Risk" in my library.



Technorati Tags:

Monday, April 13, 2009

Google Insights for Search

Google recently released a new service called Google Insights for Search. (Click on the title of this blog post to be taken to the announcement on the Google Research blog). Once you go to the site (http://www.google.com/insights/search/#) it will subsequently show up as an application in your Google Account. There is a related .pdf written by Hal Varian and Hyunyong Choi that details how to use this new and interesting tool.

In the Draft Specification we have two modules that are somewhat related to this product. Both the Analytics & Statistics Module and the Performance Evaluation Module are user defined tools that allow detailed analysis of the data in People, Ideas & Objects. The data is the key attribute to these modules. With the People, Ideas & Objects Technical Vision expecting the data volumes to explode through IPv6, Java and Wireless access to the data in a known format. Data analysis tools like Analytics & Statistics and Performance Evaluation Modules will be the key to obtaining value from it. That is essentially what Google has done with the Google Trends data, published it through a known API and interface.

The purpose in allowing users to access formatted data is the key to the value in using the tools and understanding the data. How many firms have data scattered across many departments and on individuals machine? How much of this data is available through known and trusted access mechanisms? What tools are available for people to interpret and analyze the data. We have the opportunity to embed the R application into the Modules and other opportunities.

I highly recommend downloading the R application, the .pdf from Google on how to use Google Insight for Search and Google Trends data. Experimentation with this opportunity will provide you with the beginning of understanding the opportunities we will make available to our users. And please, join me here .

Technorati Tags:

Thursday, October 02, 2008

Project Charter - Preliminary Specification

Based on the Draft Specification, this community will complete the following tasks in publishing the Preliminary Specification. Ideally the number of members of the community defining the Preliminary Specification should be up to 100 individuals. They will have a broad background in oil and gas in all professions, geographic regions and organizations, including service industry businesses.

The most important attribute of these people is they should be leaders in their field and well versed in the oil and gas industry. These People are accessing this work on the basis that their leadership skills will provide a sound Preliminary Specification, and develop a cornerstone of the "Community of Independent Service Providers" offering to the producers. They are expected to develop the Preliminary Specification on the basis where their cost of time is incurred on their behalf as an investment in defining and developing that service offering.

This investment by the leadership of "Community of Independent Service Providers" is the basis of the beginnings of these organizations. They will be able to provide users (upon approval) of their organizations to
this software and service offering. Where the People, Ideas & Objects and producer companies are their clients. The costs of Users in the Detailed and Final Specification will be fully funded before any work is started.

Establish deadlines for the Preliminary, Detailed and Final Specifications of the People, Ideas & Objects application modules. I propose the following on the basis that our $180,000 budget is raised through donations from producers. (Proposed time-lines.)
  • Draft Specification                 Published August 1, 2008.
  • Preliminary Specification        Publish May 31, 2009
  • Detailed Specification            Publish November 30, 2009 (Commence Software Development)
  • Final Specification.                Publish April 30, 2010
General overall task of the Preliminary Specification are to define the scope of the People, Ideas & Objects application modules.
  • Geographic Scope.
    • Minimum Scope: is established as North America and the North Sea.
      • What other jurisdictions regulatory and compliance requirements fit within the minimum scope.
    • Maximum
      Scope: is established as all jurisdictions that produce oil and gas and are willing to support their regulatory and compliance development costs for their jurisdiction.
  • Scope of Functionality.
    • What priorities and dependencies exist in bringing the application to market in its minimum form. (This is an iterative software development application.)
    • What "scale" related issues will this application require in terms of
      • Technology
      • Architecture,
      • User capability,
      • Community development etc.
    • Define in detail the functionality necessary for Release Candidate v 1.0.
  • Organizational Requirements.
    • Define how the People, Ideas & Objects communities are supported in their regions.
      • What unique strengths and advantages does the firm and development provide.
      • What are the weaknesses and threats of this development and firm.
    • Define the expectations and demands of the producers in developing this application.
      • Refine People, Ideas & Objects Business Model.
    • Define roles and responsibilities of the Users in the development of this software.
      • Who holds authority, whom reports to whom. What "hierarchy" is needed to organize the Users.
    • Define the requirements in terms of policies and procedures of the "Community of Independent Service Providers" qualifications and capabilities necessary to support the software in the producer firm. 
  • Process Definitions.
    • Define and map all processes in draft.
  • Algorithm Research.
      • We have some algorithms in the modules that are state of the art. Particularly in the Accounting Voucher and Partnership Accounting.We need to establish a group that is focusing on those algorithms development. This Specification should detail the needs of that group.

Project Deliverables

This is a creative process in which these Users will be able to fill in areas where the Draft Specification has missing components. The Users are not limited in any fashion in terms of the numbers of modules and
the areas they need to develop.

Detailed tactical and strategic analysis supporting the commercialization of "Release Candidate 1.0".

Preliminary Unified Modeling Language (UML) based on analysis conducted. (Please note UML is also a deliverable in both the Detailed and Final Specifications. Users will have access to the Petroleum Producer Data Model (PPDM) operating on Sun Microsystems Project Hydrazine.

Please note I have purposely kept this Preliminary Specification detail as broad as I can. The People who are this community should exercise the judgment and authority necessary to further define the Preliminary Specification as they require. 

Technorati Tags:

Saturday, June 16, 2007

Markets and the Primal-Dual Paradigm

Another Google Video presentation.

The title of this presentation is in line with the current research being done on this blog. Markets in particular are the area that is primarily "new" in determining the boundaries of the oil and gas firm and the markets it operates in. This speaker is well regraded in his work, yet very nervous in his presentation. Although he jumps around a fair amount, I think the areas that he is discussing is something of value in opening the discussion of the algorithms that are needed to be built here. Particularly when we move forward with the software developments.

The topic of discussion is markets, from the context of the algorithms that are used to determine and maintain "General Equilibrium". Much of his work seems to be based on the Nobel Prize winning economists Kenneth Arrow of Stanford University, and Gerard Debreu (deceased) and their 1954 Arrow-Debreu Theorem. The algorithms that will be used in this development will be complex. The complexity is nothing short of what appears like a very large "Manhattan Styled" project. But with the computational power available today, I am certain that it is only a matter of time, and a bit of money too.

The area of research that this presentation opens up is the area of the algorithms for this application. And I would suggest that it does not necessarily need to be discussed extensively at this point, only to be noted and added to the long list of work that needs to be done. The presenter of this material felt that their was a current revolution in the definition of markets. Noting that Google's AdWords was an algorithm of this new revolutionary area, where new "Market Models" are emerging.

If I were to summarize the one hour talk, I would present it in this introductory, yet pertinent, formula.

Total utility of a bundle of goods = equilibrium prices = each buyer gets an optimal bundle of goods + no deficiency or surplus of any good.

A formula that provides an initial understanding of the direction of the work we will need to eventually do in algorithms.

Technorati Tags: , , ,