Dr. Parikshit N.

Mahalle
Dean R&D
VIT, Pune
Agenda
 Problem Solving
 Need for Mathematical Modeling
 Steps of Mathematical Modeling
 Modeling Methods
Problem Solving
 Problem Solving means engaging in a task for which
the solution method is not known in advance.
 By learning problem solving, students should acquire
ways of thinking, habits of persistence and curiosity,
and confidence in unfamiliar situations that will serve
them well outside the classroom
 Problem solving is not knew , it is an integral part of
all mathematics learning
 George Polya (1887 – 1985) , famous mathematics
educators believed that the skill of problem solving
could and should be taught – it is not something that
you are born with.
Four Principles - steps
Polya identifies four principles (steps)that form the basis for any serious
attempt at problem solving
1. Understand the problem – what is the problem ? If you do not know
what it is , how can you solve it?
2. Devise a plan – Design a strategy , there are many known strategies
such as guess and check , or solve a simple problem or divide problem
into subproblems. Use one of them or come up with a new strategy or
action plan
3. Carry out the plan – Execute or implement the plan , - Be patient –
most problems are not solved quickly nor on the first attempt , · If a
plan does not work immediately, be persistent
Do not let yourself get discouraged . · If one strategy isn’t working, try
a different one
4. Look back (reflect) – Is there a better way ? Have I solved it
completely or there is still need for refinements
Problem definition phase
1. Understand the problem and define the problem -
 To understand the problem try to state the problem.
 Write it down, draw a picture or sketch.
 Written statement helps in focusing, avoids loosing track
 Statement need to be precise
 Do not try to solve problem before you have described to
yourself what the problem is
 Many times in a hurry to solve the problem, you get started
without understanding what the problem is and eventually
end up solving a wrong problem.
Solution Designing phase
 There are always many ways of solving a problem.
 it is difficult to decide where to start.
 Here we need to use some tried and tested methods of
problem solving.
 Some strategies, that had given fruitful solutions earlier,
can be used.
 The strategy helps in rearranging the tasks in proper
sequence, associating tools with tasks.
 Your choice of strategy may lead to knowledge of new tools
which you had not earlier envisaged. Your knowledge of
tools can also affect the strategy you tend to choose.
Implementation phase
 Implementation requires good mastery over the tools
to be used in solving the problem.
 Gathering the knowledge of tools is essential for
problem solving but this does not make you a Problem
Solver.
 Sometimes we have the action plan ready and know
the tools required for it but are unable to use them.
 We need to master tools in anticipation that we may
need them in some futuristic problem Solving effort.
Problem Solving phases

Gather
Knowledge
of Tools
Understand
Implement
& Define Refinement
ation
the
problem Design plan
of action
Refinement Phase
 Skilled problem solver will get it right in the first attempt is a
myth.
 Reflect on the previous steps critically.
 Whether the problem is defined properly? Is the problem
specification complete? You solution only partly solves the
problem, only tackles specific cases but is not taking care of all
possibilities.
 Are there better tools? Does the tools have better capabilities?
Using the new tools the quality of solution could be improved.
 Are there better designs possible for the problem? Better in terms
of cost saving. Will the new action plan result in faster, efficient
good quality solution?
 Is the implementation done properly? Are there bugs, errors?
Computational problems
 Computers can be used to solve computational problems.
 These problems intend to compute certain expected values for an
appropriate set of input values
 Computers are great data manipulators, data processors. Data
can be anything such as numbers, words, images etc.. Computers
can perform a variety of operations on data and at high speed
 An algorithm is a computer science term for a strategy or plan of
action.
 An algorithm is a step by step procedure that takes a value or set
of values as input and produces expected value or set of values as
output. The action plan guaranties the correctness or
appropriateness of the solution.
Solution Designing Phase
 The mental block is created and to overcome it you
jump to the next phase of implementing the solution
without preparing the action plan.
 Mathematics provides a rich source of tools for
problem solving
 It would be advisable to use such tried and tested
methods
 Mathematics can only be applied to mathematical
problems
 Mathematical modeling is a way of formulating a given
problem in a Mathematical form
Understanding the problem
 Solution resides in the problem
 Mathematical Modeling helps us in better
understanding the problem
 Mathematics asks for preciseness – Reduce the
ambiguities
 Modeling gives precision and direction for problem
solution
 Enables a thorough understanding of the system
Steps in Mathematical Modeling
Real world Problem

Reduced Real Problem

Mathematical Model

Analyze Test

Computer Model
Step 1 – Reducing the Problem
 Real world problem is complex, imprecise, varied
Identification
 Decide the scope of the problem
 Draw clearly what forms the part of the system and
what is outside, that is part of the environment
Environment

System
Idealization & Approximation
 Make assumptions – Precise assumptions helps in later
conversion from model to mathematical form
 State assumptions clearly and concisely
 Consider ideal situations, They are easy to handle
 Exceptional scenarios can be approximated to ideal situations
 For large complex problems several models can be derived each
with different set of assumptions
 Model need not be comprehensive but simple
 Simplify until tractable.
 Modeling concerns retaining just enough features to make the
model useful but not so many as too make it more expensive to
compute than necessary to get out the desirable information.
Step 2 – Constructing the Model
 The real entities and processes are to be replaced by
mathematical symbols , relations (sets, functions,
equations, etc.) and mathematical operations
 There may be several mathematical models for the
same real situation
 There may not be a "best" model . One can be ‘better’
than the other in certain situations
Analyzing the Model
 Once the problem is formulated into a mathematical
form, a rich set of tools can be applied to solve the
problem.
 Several tried and tested methods and techniques are
available at your disposal
 The results can be further used to improve the model
Testing the Model
 Determine whether the model gives predictions
consistent with observations.
 Strong disagreement for certain situation can be
because of omissions or assumptions made in
reduction phase
Modeling methods
 Understand the available tools before you make a
choice
 Master the available tools before you use them
 Look at how others model similar situations; so that
you can adapt their models to your situation.
Modeling with Difference
equations
 If the current value of the variable is an then the predicted
value of the variable will be an+1 where
an+1 = αan + β
 In some instances, to predict a new value goes back further
than the current value, e.g.,
an+1 = an + an−1
 Finally, it may be that the form of the difference equations
are unknown and the model must be
Written
an+1 = f(an; an−1; an−M−1)
 Determining the nature of f and the step M is at the heart
of model formulation with difference equations.
Modeling with Differential
equations
 Differential equations use time as a continuous variable
 One of the simplest differential equations is also an
extremely important model, i.e.,
dx/dt = x
 In words, the rate of change of the quantity x depends on
the amount of the quantity.
 It may be that the instantaneous rate of change at time t is
actually a function of a previous time, i.e.,
dx /dt = f(x(t)) + g(x(t −τ ))
 This is referred to as a delay differential equation.
Partial differential equations
 Many situations arise in practice where the number of
independent variables is larger than two.
 For spatio-temporal models we might have time and
space
Linear programming Models
 Many problems are optimization problems where the
goal is to maximize profit or to minimize cost
 It is to achieved under certain constraints
 If the objective function as well as the equations that
define the constraint set are linear, than the
optimization problem is called a linear programming
problem.
 There are several problem scenarios that can be posed
as LP problems
 Assignment problem, Transportation problem,
Transhipment problem
Data Modeling - Regression
 Data is available from observations of the process
 A line or curve can be fitted to the data to get the
empirical relationship
 A function representing the data is formulated such as

 The parameters involved are derived from observed

data
Network models
 Network models are an extremely important category
of mathematical Models
 The largest mathematical programs that are regularly
solved in practice, e.g. airline crew scheduling
problems, are usually network problems.
 Network models are created from two major building
blocks: arcs (sometimes called edges), which are
connecting lines, and nodes, which are the connecting
points for the arcs.
 Additionally, a network is a graph (or more commonly
a digraph) in which the arcs have an associated flow
Reverse Engineering
Problem

Known Mathematical Mathematical Model

Model

Known Programming Programming Model

Solutions
Example
 An oil refinery produces two products: jet fuel and gasoline. The profit
for the refinery is
 $0.10 per barrel for jet fuel and $0.20 per barrel for gasoline. The
following conditions must be met.

 1. Only 10,000 barrels of crude oil are available for processing.

 2. The refinery has a government contract to produce at least 1,000
barrels of jet fuel
 3. The refinery has a private contract to produce at least 2,000 barrels of
 gasoline.
 4. Both products are shipped in trucks, the delivery capacity of the
truck fleet is 180,000 barrel-miles.
 5. The jet fuel is delivered to an airfield 10 miles from the refinery.
 6. The gasoline is transported 30 miles to the distributor.

 How much of each product should be produced for maximum profit?

Example
 Let x represent the number of barrels of jet fuel and y
represent the number of barrels of
 gasoline. Then the profit function is p(x,y) = 0.10x +
0.20y. The constraints are as follows…

 x + y <= 10,000
 x >= 1000
 y >= 2000
 10x + 30y <= 180,000
References
 MATHEMATICAL MODELING A Comprehensive
Introduction By Michael Kirby and Gerhard
Dangelmayr
 Mathematical Modeling by Arnold Neumaier
 Any Doubt ?

Please feel free to write to

me:
aalborg.pnm@gmail.com
98 22 416 316

Thank You !!!