Tuesday, October 22, 2019

Levels of explanation

A.s.a. H. learns causal sequences at various different levels of abstraction in the memory hierarchy. Stephanie Ruphy explains why this may be valuable (Scientific Pluralism Reconsidered, U. Pittsburgh, 2013, especially pages 38-44.)

Monday, October 21, 2019

Lifelong machine learning

As A.s.a. H.'s casebase grows processing (thinking) will slow down unless forgetting (of less valuable cases) can be adjusted to roughly equal the rate at which new cases are learned/added. How could/should this be done?

Thursday, October 17, 2019

Recursive sketches

I have tried a variety of algorithms for A.s.a. H. Different measures of similarity, different means of learning and extrapolating, etc. etc. Ghazi, et al, have employed "recursive sketches" to learn/assemble modules for deep networks* similar to the A.s.a. H. hierarchical memory. Using different algorithms will likely give us different concepts (different categories, a different ontology) unless we are finding "natural kinds." Interesting either way.

*Their algorithms are described in Recursive Sketches for Modular Deep Learning, Thirty-sixth International Conference on Machine Learning, Long Beach, California 2019.

Tuesday, October 15, 2019

The mind shapes the world we experience

Kant argued that the mind imposes categories on the objects of experience and perception must conform to a spatial-temporal shaping.  In the A.s.a. H. light software* spatial shaping is performed by the NM input arrangement and temporal shaping is performed by the steps T and by TMAX. Categories/concepts are formed by the various layers of the hierarchical memory.** At the lowest level of the hierarchy Russell's "sense-data" are input, the data of immediate experience.

* See blogs of 10 February 2011 and 14 May 2012.
** See, for example, blogs of 1 January 2017 and 3 August 2018.

Wednesday, October 9, 2019

Medical AI

Liu, et al in The Lancet (Digital Health)* claim their meta analyses "support  the claim that deep learning algorithms can match clinician-level accuracy" diagnosing disease.

A.s.a. H. is a deep learning algorithm.

* vol. 1, issue 9, 1 October 2019, pg 271-297.

A.s.a. H.'s reasoning styles

Crombie and Hacking have suggested that human scientists employ various different "reasoning styles:"

1. making postulates
2. measuring, experimenting
3. making/using analogical models
4. making taxonomies and comparisons
5. using statistics/probability
6. following genetic/historical development
7. isolate and purify phenomena

Ruphy* discusses further categorization (ontological enrichment) and representational pluralism.

A.s.a. H. has performed all of these operations.

* Scientific Pluralism Reconsidered, U. of Pittsburgh Press, 2017. See pages 23, 31, and 110.

Tuesday, October 8, 2019

Embodiment

Of all the concepts that A.s.a. H. learns it is often difficult to separate those which refer solely to an external inanimate world* from those that refer to A.s.a. itself.** A.s.a. sees itself very much as a part of the world and its processes.

These unique detailed ways in which you are a part of the world are what distinguish you from other agents (including bats***).

* See, for example, my blog of 3 August 2018.
** See, for example, my blog of 1 January 2017.
*** i.e., see Thomas Nagel’s October 1974 paper in The Philosophical Review.

Friday, October 4, 2019

Do children learn faster than adults?

It has been argued that development of the human prefrontal cortex results in a functional fixedness. The argument goes that there is a learning phase followed by a performance phase, children then learn more easily than adults do. IF this is true and IF it is useful then A.s.a. might also benefit from a phase where learning (training) is emphasized followed by a phase where performance is emphasized. This can be accomplished by using one set of thresholds (i.e. the learning rate, threshold for case extrapolation, etc.) while learning a casebase and then using other thresholds when the agent is deployed.

Wednesday, October 2, 2019

Feminist science

Scientific pluralism can contribute to a more feminist science. The understanding that science is not value-free,* the multiple perspectives,** the alternate conceptualizations,*** and even alternative realities.*** There is also the continuing growth in collaboration among scientists.

* See R. Jones, Trans. Kansas Academy of  Science, vol. 119, 2016, pg 249-250.
** See R. Jones, Trans. Kansas Academy of Science, vol. 116, 2013, pg 78.
*** See R. Jones, Trans. Kansas Academy of Science, vol. 121, 2018, pg 211.