The document discusses artificial neural networks (ANNs). It provides an overview of ANNs, including their biological inspiration from neurons in the brain, their composition of interconnected processing elements called neurons, and how they are configured for applications like pattern recognition. The document also covers different types of ANNs, their computational power, capacity for learning, convergence abilities, and use for generalization. Examples are given of ANN applications in various business domains like marketing, sales forecasting, finance, insurance, and telecommunications. Risks of ANNs discussed include needing a large and diverse training set, overfitting data, and high hardware resource requirements. A hybrid symbolic-neural network approach is also mentioned.

1. Presented By:
Akif
Burhan Muzafar
Junaid Hassan
Umar manzoor
Muneer AH
Saleem zahoor

2.  An Artificial Neural Network (ANN) is an
information processing paradigm that is
inspired by biological nervous systems.
 It is composed of a large number of highly
interconnected processing elements called
neurons.
 An ANN is configured for a specific
application, such as pattern recognition or
data classification

3.  The neuron receives signals from other neurons,
collects the input signals, and transforms the
collected input signal.
 Human information processing takes place
through the interaction of many billions of
neurons connected to each other, each
sending excitatory or inhibitory signals to
other neurons (excite in positive/suppress in
negative)

4. Artificial neural networks (ANNs) are biologically
inspired computer programs designed to simulate the way
in which the human brain processes information. ANNs
gather their knowledge by detecting the patterns and
relationships in data and learn (or are trained) through
experience.

5. Types of artificial neural networks
 Artificial neural network types vary from those with
only one or two layers of single direction logic, to
complicated multi–input many directional feedback
loops and layers. On the whole, these systems use
algorithms in their programming to determine control
and organization of their functions.

6.  Computational power
The multi-layer perception (MLP) is a universal
function approximate, as proven by the universa
approximation theorem. However, the proof is not
constructive regarding the number of neurons required
or the settings of the weights. Work by Hava
Siegelmann and Eduardo D. Sontag has provided a
proof that a specific recurrent architecture with rational
valued weights (as opposed to full precision real
number-valued weights) has the full power of
a Universal Turing Machine using a finite number of
neurons and standard linear connections. Further, it has
been shown that the use of irrational values for weights
results in a machine with super-Turing power.

7.  Capacity
Artificial neural network models have a property called
'capacity', which roughly corresponds to their ability to
model any given function. It is related to the amount of
information that can be stored in the network and to the
notion of complexity.
 Convergence
Nothing can be said in general about convergence since it
depends on a number of factors. Firstly, there may exist
many local minima. This depends on the cost function and
the model. Secondly, the optimization method used might
not be guaranteed to converge when far away from a local
minimum. Thirdly, for a very large amount of data or
parameters, some methods become impractical. In general,
it has been found that theoretical guarantees regarding
convergence are an unreliable guide to practical application.

8.  Generalization and statistics
In applications where the goal is to create a system that generalizes
well in unseen examples, the problem of over-training has emerged.
This arises in convoluted or over-specified systems when the capacity
of the network significantly exceeds the needed free parameters. There
are two schools of thought for avoiding this problem: The first is to
use cross-validation and similar techniques to check for the presence of
overtraining and optimally select hyperparameters such as to minimize
the generalization error. The second is to use some form
of regularization This is a concept that emerges naturally in a
probabilistic (Bayesian) framework, where the regularization can be
performed by selecting a larger prior probability over simpler models;
but also in statistical learning theory, where the goal is to minimize
over two quantities: the 'empirical risk' and the 'structural risk', which
roughly corresponds to the error over the training set and the predicted
error in unseen data due to overfitting.


10. Over the last decade, neural networks have found
application across a wide range of areas from business,
commerce and industry. Following an overview is provided
of the kinds of business problems to which neural networks
are suited, with a brief discussion of some of the reported
studies Relevant to each area.

11.  The goal of modern marketing exercises is to
identify customers who are likely to respond
positively to a product, and to target any
advertising or solicitation towards these
customers. Target marketing involves market
segmentation, whereby the market is divided into
distinct groups of customers with very different
consumer behavior. Market segmentation can be
achieved using neural networks by segmenting
customers according to basic characteristics
including demographics, socio-economic status,
geographic location, purchase patterns, and
attitude towards a product.

12.  Businesses often need to forecast sales to make
decisions about inventory, sta$ng levels, and
pricing. Neural networks have had great success
at sales forecasting, due to their ability to
simultaneously consider multiple variables such
as market demand for a product, consumers'
disposable income, the size of the population,
the price of the product, and the price of
complementary products. Forecasting of sales in
supermarkets and wholesale suppliers has been
studied and the results have been shown to
perform well when compared to traditional
statistical techniques like regression, and human
experts.

13.  One of the main areas of banking and "nance
that has been affected by neural networks is
trading and financial forecasting. Neural
networks have been applied successfully to
problems like derivative securities pricing and
hedging , futures price forecasting, exchange
rate forecasting and stock performance and
selection prediction.

14.  There are many areas of the insurance industry
which can benefit from neural networks. Policy
holders can be segmented into groups based
upon their behaviors, which can help to
determine effective premium pricing. Prediction
of claim frequency and claim cost can also help
to set premiums, as well as find an acceptable
mix or portfolio of policy holders characteristics.
The insurance industry, like the banking and
finance sectors, is constantly aware of the need
to detect fraud, and neural networks can be
trained to learn to detect fraudulent claims or
unusual circumstances.

15.  Like other competitive retail industries, the
telecommunications industry is concerned with the concepts
of churn (when a customer joins a competitor) and win back
(when an ex-customer returns). Neural Technologies Inc., is
a UK-based company which has marketed a product called
DA Churn Manager. Specifically tailored to the
telecommunications industry, this product uses a series of
neural networks to: analyze customer and call data; predict
if, when and why a customer is likely to churn; predict the
elects of forthcoming promotional strategies; and
interrogate the data to find the most profitable customers.
Telecommunications companies are also concerned with
product sales, since the more reliant a customer becomes on
certain products

16. Models of
ARTIFICIAL NEURAL
NETWORK

17. Four parts of a typical
nerve cell : -
 DENDRITES: Accepts the
inputs
 SOMA : Process the
inputs
 AXON : Turns the
processed inputs into
outputs.
 SYNAPSES : The
electrochemical
contact between the
neurons.

18.  Inputs to the network are
represented by the
mathematical symbol, xn
 Each of these inputs are
multiplied by a connection
weight
 These products are simply
summed, fed through the
transfer function, f( ) to
generate a result and then
output.
f
w1
w2
xn
x2
x1
wn
f(w1 x1 + ……+ w

19. output layer
connections
Input layer
Hidden layers
Neural
network
Including
connections
(called
weights)
between
neuron
Com
pare
Actual
output
Desired
output
Input
output
Figure showing adjust of
neural network
: artificial neural network model
CONTD

20. The neural network in
which every node is
connected to every other
nodes, and these
connections may be
either excitatory
(positive weights),
inhibitory (negative
weights), or irrelevant
(almost zero weights).
These are networks in
which nodes are
partitioned into subsets
called layers, with no
connections from layer j
to k if j > k.
Input node
Input node
output node
output node
Hidden node
Layer 1 Layer2
Layer0
(Input layer) (Output layer)

21.  Neurons in an animal’s brain are “hard
wired”. It is equally obvious that animals,
especially higher order animals, learn as
they grow.
 How does this learning occur?
 What are possible mathematical models of
learning?
 In artificial neural networks, learning refers
to the method of modifying the weights of
connections between the nodes of a
specified network.
 The learning ability of a neural network is
determined by its architecture and by the
algorithmic method chosen for training.

22. TRAINING RISK
HARDWARE RISK
HYBRID APPROACH

23.  A common risk of neural networks, particularly in
robotics, is that they require a large diversity of
training for real-world operation. This is not
surprising, since any learning machine needs
sufficient representative examples in order to capture
the underlying structure that allows it to generalize to
new cases. Dean Pomerleau, in his research
presented in the paper "Knowledge-based Training of
Artificial Neural Networks for Autonomous Robot
Driving," uses a neural network to train a robotic
vehicle to drive on multiple types of roads (single
lane, multi-lane, dirt, etc.). A large amount of his
research is devoted to extrapolating multiple training
scenarios from a single training experience, and

24.  Preserving past training diversity so that the
system does not become over trained (if, for
example, it is presented with a series of right turns –
it should not learn to always turn right). These
issues are common in neural networks that must
decide from amongst a wide variety of responses,
but can be dealt with in several ways, for example
by randomly shuffling the training examples, by
using a numerical optimization algorithm that does
not take too large steps when changing the network
connections following an example, or by grouping
examples in so-called mini-batches.

25.  To implement large and effective software neural networks,
considerable processing and storage resources need to be
committed. While the brain has hardware tailored to the task of
processing signals through a graph of neurons, simulating even a
most simplified form on Von Neumann technology may compel a
neural network designer to fill many millions of database rows
for its connections – which can consume vast amounts of
computer memory and hard disk space. Furthermore, the
designer of neural network systems will often need to simulate
the transmission of signals through many of these connections
and their associated neurons – which must often be matched with
incredible amounts of CPU processing power and time. While
neural networks often yield effective programs, they too often do
so at the cost of efficiency (they tend to consume considerable
amounts of time and money).

26.  Some other risks come from advocates of
hybrid models (combining neural networks
and symbolic approaches), who believe that
the intermix of these two approaches can
better capture the mechanisms of the human
mind.