DebKR

To the Stars

Artificial Intelligence

Deep Learning

Deep Learning

By

I still haven’t had time to check out Google’s open source deep learning technologies but I know I need to get up to speed on this. They’ve now released Tensor2Tensor (T2T) to enable use of their machine intelligence library, TensorFlow, for training deep learning mobiles. Useful for translation, parsing, image captioning and lots of other great stuff, T2T comes with a library of research data sets and models.

Get T2T on GitHub: https://github.com/tensorflow/tensor2tensor

Original news article picked up from SDTimes: http://sdtimes.com/tensor2tensor-xebialabs-devops-syncsort-sdtimes-news-digest/

 

Effective Study Skills 2

Effective Study Skills 2

By

This post is a continuation of my summarising of learnings from Dr Oakley’s MOOC on Learning How to Learn (see References below) together with thoughts and reflections on that course in relation to other courses currently being studied. Related posts can be found under the tag Learning About Learning.

 

—

The Illusion of Competence

the employing of ineffective study methods which fool the mind into thinking it’s learned the information, when in fact it hasn’t

Overlearning

where the information is engrained in your mind but in unhelpful & inefficient ways, making reaccessing that information more difficult

Interleaving

the strategy of providing variety into your study routine – different subjects, different learning modalities, and so on – to make it more interesting & engaging which helps with information retention

 

References:

Ison, R. (2010) Systems Practice: How To Act In A Climate-Change World, Springer, London.

Learning How to Learn: Powerful mental tools to help you master tough subjects, University of California, San Diego. Taught by: Dr Barbara Oakley & Dr Terrence Sejnowski. Available at: Coursera.org.

How We Find Out Things

How We Find Out Things

By

How We Find Out Things is important. Intelligence is the ability to find things out about a situation or subject, and to apply what we have previously found out to a new or novel situations or subjects.

Machine Learning does this through methods of Computational Statistics. It is powerful, but we still need to be there applying our own Intelligence & Sense-making abilities to ensure the results are logical & consistent with what we would expect.

As Ramon y Cajal, Spanish biologist and father of modern neuroscience, noted in his excellent book of advice for novice scientific researchers (1999), the key stages of a scientific investigation are (1) observation, (2) experimentation, (3) working hypotheses, and (4) proof. [Read more…] about How We Find Out Things

Effective Study Skills

Effective Study Skills

By

This post summarises learnings from Dr Oakley’s MOOC on learning How to Learn (see References below for link) together with my own thoughts and reflections on that course together with my Systems Theory post-grad course currently being studied with the Open University and other current studies. Related posts can be found under the tag Learning About Learning.

In particular, I have taken time out from that post-grad module due to feeling really confused, fed up, bored and generally utterly demoralised by my continually unsuccessful attempts to get on with studying that module. Rather than beat myself up about it, I decided to just give myself some time off to go and study a different (albeit, connected) subject. Hopefully it will free up some mind-space and sparks some useful ideas to go towards my next assignment.

Chunking as a Learning Method

Chunking is studying information in small, accessible packages. These chunks are more easily assimilated, since the mind is better able to unite and fit them together into a unified whole. Chunking has been described as a ‘mental leap’ – taking pieces of a jigsaw puzzle and fitting them together to make a whole picture. [Read more…] about Effective Study Skills

Perception

Perception

By

Asked question: What else do I need [to study/learn] for AI?
Answer on waking: Perception

Thinking on this a little, I realise perception must be key bcs this is the route/mechanism by (or, gateway through) which one experiences the world (= other) & from this one can begin to sort, categorise, pattern-seek & understand.

A realisation stemming from this is: without perception there can be no cognition (which I say must also be vital for intelligence) bcs perception creates the stimuli which ‘feed’ the cognitive processes.

Another key realisation is: ‘discrimination’ (used in its true sense) is the basis of intelligence. If you can’t perceive difference then you can’t create categories thus you can’t compute anything about either ‘self wrt other’ or ‘other wrt self’ or even ‘other1 wrt other2’, & so on. [Read more…] about Perception

MLND Project: Model Evaluation & Validation

MLND Project: Model Evaluation & Validation

By

3e Project: Model evaluation & validation

Project details

DESCRIPTION:

  • use Boston house price dataset to PREDICT selling price of a new/unseen home

PROCESS:

  1. EXPLORE data > obtain important FEATURES & DESCRIPTIVE statistics abt data
  2. Properly SPLIT dataset into TRAINING & TEST datasets
  3. DETERMINE suitable PERFORMANCE METRIC for evaluating the problem
  4. ANALYSE performance graphs for learning algorithm over varying TRAINING SET SIZES & with varying number of PARAMETERS
  5. CHOOSE OPTIMAL MODEL that best generalises unseen data
  6. TEST chosen optimal model on a NEW SAMPLE & COMPARE PREDICTED selling price to ACTUAL statistics

[Read more…] about MLND Project: Model Evaluation & Validation