DWH News

Automated code analysis relieves Developers from having to perform line-by-line inspections during peer code reviews, freeing them to focus on higher-level analyses that require human intelligence. With automated checking for compliance to the team's and organization's development policies, the team can begin peer code reviews by discussing interesting findings from the automated code analysis results, then move on to examining high-level design, algorithmic, and implementation issues.

A continuous, automated process centralizes management of pattern-based static analysis, data flow static analysis, and code metric calculation. Integrating static analysis into existing development workflows is key to establishing a practical and sustainable process. Moreover, to ensure that the resulting process remains on track, reports provide managers visibility into policy adherence and alert them to areas where process improvement might be valuable.

I argue that good research design and modelling can take advantage of, and incorporate, the phenomena of "predictable irrationality", and canvas some of the issues raised in a recent whitepaper and book.

A new forecasting competition, with the emphasis on "computational intelligence" forecasting methods, and an interesting data set .. cash withdrawals from 111 ATMs

YAML, or the YAML concept, may offer a way forward for text based literate and agile data entry.

I consider an example of "Crunching" before "SuperCrunching" .. what appears to be a reasonably well performed analysis of "stockmarket investing" based on historical data. Some niggles surface.

modelling individual decision processes, and embracing verbal explanations .. I buy a new mower

In which I declare PR dead, and muse on how to get an honest estimate of SERPs

I needed some PSO ("Particle Swarm Optimization") for a Visualization App, and I thought that I could get a QUICK FIX by PORTING some LEGACY CODE written in a DEAD LANGUAGE, only it didn't work out that way .. AND IT NEVER does: elegant literate numerate rewrites rule, OK?

a positive post, in which I express appropriate gratitude to Microsoft that FTP can now transfer data around the web (like it used to) - if you get the fix -, that Excel can now multiply numbers and get the same answers as a pocket calculator - if you get the fix -, and that Hungarian Notation is no longer de-rigueur (you will have to supply your own fix for that).

thoughts on using modelling to form "reasonable expectations" of possible outcomes of marketing actions; data driven estimates of limits on influence. So we don't leave ourselves quite so open to wishful thinking, or to the post-hoc comeback of "well, that was bloody stupid"

a couple of blogs that I had missed, with some interesting articles on subjects as diverse as human-machine partnerships in decision making, the proper assessment of risk and the dependence of test outcomes on the background population, a meld of a GA with a Random Forest approach and more approaches to modelling decision making. A bit of a grab bag really. Nothing of mine, just stuff that caught my eye.

a bunch of reasons why I decided to resume blogging .. or is that justifications for my decision?