Behavioral finance FAQ / Glossary (Neural net)

A    B    C    D    E    F    G-H    I-L    M    N-O    P-Q    R    S    T-U    V-Z

Full list

This is a separate page of the N-O section of the Glossary

 

Dates of related message(s) in the
Behavioral-Finance group (*):
Year/month, d: developed / discussed,
i: incidental

Neural / neuronal net

00/6i,7i,9i,10i,12d - 01/11i -
05/1i,4i + see softcomputing, AI,

genetic algorithm, agent-based
model, system trading, quant

The mind in the software.

Definition

Neural nets (or artificial neural network) are computer
hardware or software
that try to imitate the brain neuron connections, or
at least the human reasoning and learning process (with trial and
errors, associations of ideas...).


The purpose, like in most computerized decision processes (see
genetic algorithm) is is to gain efficiency, and at the same time to
keep a human control in the way the process is designed and on
how ir works in practice


What kind of output?

Neuronal software are used to :

* make speedily some routine decisions, without human
   intervention, even in complex situations (*) with various
   (or fast moving) factors
involved.

* make various types of simulations and previsions, i
  
ncluding economic and financial "war games".

(*) but maybe not, as seen below, when facing totally
     new phenomena.

A lot of research is done to improve them.

Coupled with other "soft computing" (see that phrase) tools, they are the
bases of sophisticated mathematical modeling that aims at "artificial
intelligence".

The man - machine contest


Here comes the philosophical questions:

OK, computers can ape us, emulate some of our reasoning,
and do a few things faster than us.

But can they do it better than us?

(see artificial intelligence, transhumanism)
And can they,
as emotionless creatures, avoid some of our
biases ?
(see behavioral biases)

Can we rest on the beach while those workhorses
do the hard decision job

For routine / autopilot situations,

yes, robots can be superior to us

But better watch them, as near-automatic moves (see
heuristics) also can become risky biases.

Even more when

* "rare events" (see that ohrase) strike,

* and/or creativity and vision are needed.

* to face exceptional events,
* to take creative initiatives,
* and on situations that involve sentiments

Those areas of decisions are not really their
turf.  Until now...

What are they used for
     in
economics and finance?

Tools for market plumbers?

The use of neural software as economic / financial explanation, projection
or decision tools, is mostly in an experimental phase. But it is seen already:

In various microeconomics models

In credit analysis (by crossing data on factors that might lead to
    bankruptcy),

In quantitative financial analysis tools,

In stock trading systems (in which speed is essential, but also can
    be disastrous)

In other portfolio management systems (reallocation of asset classes)

Neuroeconomics, Neurofinance, 
   Neuroscience

See the
specific glossary page

(*) To find those messages: reach that BF group and, once there,
      1) click "messages" 2)
enter your query in "search archives".

Members of the BF Group, please
 vote on the glossary quality at
BF polls

separ

This page last update: 10/09/15  

    N-O section of the Glossary
Behavioral-Finance Gallery main page

Disclaimer / Avertissement légal