Posts

Some Thoughts on Deep Learning
Image
 Used To   He is used to the heat. He is used to being the underdog. He used to come here often. He used to be a boxer. The rake used to aerate the soil The rake is used to aerate the soil. The rake used to aerate the soil was stolen.                           The rake (that was) used to aerate the soil … The rake used to be used to aerate the soil. Making “used to” into a wordgroup immediately will lead to a lot of grief. One possible way is to have patterns of acceptance, so it is checked for the right meaning   – something like Person   IsWasVerbAuxiliary          used to                  NounPhrase Person   IsWasVerbAuxiliary          used to                 ParticipialPhrase Something           used to                  BaseForm Something           IsWasVerbAuxiliary          used to                  BaseForm Something           (that was)             used to                  Baseform             NounPhrase        VerbPhrase Something          used to
  Large Language Models Google is touting its Large Language Model, PaLM, with a claimed 540 billion nodes. It learns by reading text. A few problems. Firstly, complex text – a piece of legislation, a high value contract or a Defence specification – has its own definitions, in a glossary or declared somewhere in the document. Words or phrases defined this way are meant to override the common meanings. state of mind of a person includes:                      (a)   the knowledge, intention, opinion, suspicion, belief or purpose of the person; and                      (b)   the person’s reasons for the intention, opinion, belief or purpose.   Secondly, complex text relies heavily on bullets, with references to a particular bullet capturing the phrase or clause it links to. (b) a regular premium policy to which paragraph (a) does not apply corporate group has the meaning given by subsection 123(12).   In other words, a complex text document is its own thing, an
Image
 Training Wheels for Active Structure Taking Active Structure from knowing very little to knowing a lot. The slideshow assumes the network has definition strings from a dictionary, but so far none is built into structure PDF SlideShow
  Anti Money Laundering Using Active Structure   The term "money laundering" refers to the processes by which the illegal origins of funds are concealed, usually through interactions with banks and businesses. An estimated 2.7% of the world's global GDP is laundered annually, equating to huge government losses on a global scale. These are the funds that allow criminal and terrorist organizations to continue their operations financially unhindered. The massive problem posed by money laundering has resulted in a huge market for anti-money laundering software. Despite increasing investments in software development, an estimated 90% of money laundering still goes undetected. So what's been going wrong? Why has the existing software been performing so poorly?   Let's explore some of the approaches being used. One can take an algorithmic approach and write a program to deal with the issue. Such a program can even be quite complex and take into account a variety
Image
 The Rear Vision Mirror The Tweet by Paul H√ľnermund "Causal inference isn't a subfield of statistics. Synthetic control methods are just propensity score matching along the time dimension. G-computation is a stupid name. And DAGs are mainly good for teaching." There you have it. set me thinking. Why do people like looking in the rear vision mirror only. They obviously don’t do it when they drive a car. What about a mass of data convinces them there will be a different outcome to being splattered on the shrubbery? Yes, the picture may be clearer in the rear vision mirror, but the things you need to see to avert disaster can only be seen through the windscreen. Button-up boots is a good example. Sales were great, until someone invented a zipper – a disruptor. Instead of poring over sales data, the time would have been better spent checking relevant patent applications and anything else happening in the market (cheap imports). People who market the message – TH