The Core Aim of Mathematics Mastery

Did you know that mathematics can help a business choose how much to produce when it cannot afford to make everything? Imagine a small workshop with limited materials, limited packing space and two useful products. Every extra product earns something, but each also uses resources. The challenge is not simply to make more. It is to find the best feasible combination.

The core aim of linear programming mastery is to translate decisions into variables, write an objective function, turn practical restrictions into linear inequalities, identify a feasible region and choose a maximum or minimum value that obeys every constraint. Students also learn what corner points mean, why an attractive numerical answer can be impossible in practice, and when real-world issues such as whole-number quantities change the result. Linear programming is the mathematics of making the best choice within limits.

This article connects eduKateSG’s Inequalities, Coordinate Geometry, Simultaneous Equations and Mathematical Modelling. It complements the broader specialist guide How Mathematical Optimisation Works and the applied article Why Mathematics? | Supply Chains, Linear Programming and Delivery Planning. This page focuses specifically on learning, calculating and checking linear programming.

What Is Linear Programming?

Linear programming is a method of maximising or minimising a linear expression subject to linear constraints. The quantity we want to optimise is the objective function. The limits we must obey are the constraints. The set of all choices satisfying the constraints is the feasible region.

It is called programming in the older mathematical sense of planning and allocating resources—not because the question necessarily involves computer code.

Define Decision Variables Before Writing Formulas

Suppose a workshop makes notebooks and folders. Let:

x = number of notebooks produced;
y = number of folders produced.

Because we cannot make a negative number of products, x≥0 and y≥0. If each item must be whole, we would also require integer values. In a basic graphical linear programming exercise, however, we first solve the continuous version and then examine whether whole-number restrictions matter.

This is one place where a written sentence is worth more than rushing into equations: a clearly defined variable tells every reader what the answer will mean.

Write the Objective Function

Suppose a notebook contributes $3 toward profit and a folder contributes $2. The workshop wants to maximise total contribution:

Maximise P=3x+2y.

The objective is not automatically sales revenue. If the problem specifies profit, costs or contribution margin, use that precise quantity. Mixing revenue with profit can produce a perfectly calculated but commercially wrong decision.

Translate Material Limits Into Inequalities

Suppose each notebook requires two units of material and each folder requires one unit. The workshop has eight material units available.

The constraint is:

2x+y≤8.

Why ≤ rather than =? Because the workshop can use less than all available material. An equality would force every last unit to be used, which is not what “at most eight” means.

Translate Packing Limits Into a Second Inequality

Suppose the packing team can handle no more than six items in total, regardless of product type.

That gives:

x+y≤6.

The full model is therefore:

Maximise P=3x+2y;
subject to 2x+y≤8;
x+y≤6;
x≥0 and y≥0.

The Feasible Region Contains Every Allowed Combination

Each inequality describes a half-plane. The feasible region is where all those half-planes overlap.

In this two-variable problem, the region lies in the first quadrant, below both resource boundary lines. Any point outside the feasible region violates at least one condition.

For instance, (x,y)=(5,5) looks productive, but it uses 2(5)+5=15 material units and makes 10 items. It violates both limits. It is not a valid answer, no matter how large its objective value appears.

How to Draw the Feasible Region

A quick test point such as (0,0) can help determine which side to shade when the origin does not lie on the boundary.

Corner Points Often Locate the Optimum

For a linear objective over a nonempty bounded polygonal feasible region, at least one maximum and one minimum occur at a vertex.

That is why a standard graphical linear programming method can evaluate the objective at the feasible region’s corner points rather than test every possible interior point.

The geometric reason is that lines of constant objective value, such as 3x+2y=c, slide parallel to one another. The best line that still touches the feasible region must meet it at a boundary; a vertex is one possible point of contact.

Worked Example: Find All Feasible Corner Points

For our workshop, the feasible polygon has four vertices:

(0,0) — produce nothing;
(4,0) — material limit meets the x-axis;
(2,4) — the two resource boundary lines intersect;
(0,6) — packing limit meets the y-axis.

To find the intersection of the two resource boundaries, solve:

2x+y=8 and x+y=6.

Subtract the second equation from the first:

x=2.

Substitute into x+y=6:

y=4.

Thus the two resource limits meet at (2,4). This also shows why Simultaneous Equations is a direct prerequisite for linear programming.

Worked Example: Calculate the Best Objective Value

Evaluate P=3x+2y at all four feasible vertices:

Feasible point (x,y)	Objective calculation	Contribution P
(0,0)	3(0)+2(0)	$0
(4,0)	3(4)+2(0)	$12
(2,4)	3(2)+2(4)	$14
(0,6)	3(0)+2(6)	$12

The maximum is P=$14 at (x,y)=(2,4).

Interpretation: produce two notebooks and four folders. This uses 2(2)+4=8 material units and 2+4=6 packing slots, satisfying both capacity constraints exactly. In this particular example, the solution is already in whole numbers, so no additional integer adjustment is needed.

Why the Most Profitable Product Is Not Always the Only Choice

Notebooks contribute $3 each, more than the $2 contributed by a folder. It might therefore seem sensible to make only notebooks.

But notebooks consume twice as much material. Material is scarce here. The best solution balances contribution per item against the resources used. This is the core logic behind optimisation with constraints.

Binding and Slack Constraints

A binding constraint holds with equality at the chosen solution. A slack constraint leaves unused capacity.

At (2,4), both 2x+y=8 and x+y=6. Neither material nor packing capacity has slack. At (4,0), material is fully used but only four of six packing slots are used, leaving two units of packing slack.

Recognising binding constraints helps students explain why the optimum is where it is.

Several Best Solutions Can Exist

A linear programming problem may have more than one optimal point. For example, if an objective line is parallel to a binding edge of the feasible region, every point along that edge can give the same optimal value.

Students should not assume the answer must always be one isolated vertex. If adjacent vertices have the same optimum, the segment connecting them may contain other optimal solutions.

What Happens When No Feasible Solution Exists?

Constraints can contradict one another. For example, requiring x≥10 and x≤5 produces no feasible value.

In two variables, the feasible half-planes may have no common overlap. The model is then infeasible. There is no optimal point because there is no valid candidate at all.

An Unbounded Region Does Not Always Mean an Unbounded Optimum

A feasible region can extend indefinitely. Whether the objective also increases indefinitely depends on its direction.

For example, minimising x+y over x≥0 and y≥0 has the finite optimum 0 at (0,0), despite the feasible region being unbounded. Maximising x+y over the same region has no finite maximum.

This distinction matters because graph shape and objective direction must be interpreted together.

Integer Restrictions Change Some Answers

Ordinary linear programming lets decision variables take real values. That is appropriate for quantities such as kilograms or litres in many models, but not always for people, vehicles or individual products.

When variables must be whole numbers, the model becomes an integer programming problem. Rounding the continuous optimum is not generally a valid method: the rounded point may violate a constraint, or another feasible integer point may be better.

How Real Resource Allocation Gets More Complicated

Real operations may need to model labour shifts, minimum order sizes, equipment changes, uncertainty, delivery deadlines, nonlinear costs and dependencies between decisions.

Linear programming is valuable when a linear model is a suitable approximation. It does not mean every complicated real-world decision can be represented accurately by a few straight lines. Strong students learn to state and test assumptions.

What Is the Simplex Method?

Graphing is practical for two decision variables. Larger linear programmes may have hundreds or thousands of variables and constraints, so software algorithms become important.

The simplex method is a classic algorithm that moves among suitable basic feasible solutions in pursuit of improved objective values. Interior-point methods are another major family of optimisation algorithms.

The school-level graphical problem provides the geometric intuition behind more advanced computational methods.

Shadow Prices Explain the Value of More Resources

A shadow price describes how much the best objective value changes when the right-hand side of a constraint increases by one unit, within an appropriate range where the local optimisation structure remains stable.

In a production problem, that can help answer whether acquiring an additional unit of material or packing capacity is more valuable. This is an advanced extension, not a number to guess from the final objective alone.

Linear Programming Connects to Inequalities and Graphs

The constraints define half-planes using Inequalities. The boundaries are straight lines studied in Coordinate Geometry. Intersections require Simultaneous Equations. The entire problem requires Mathematical Modelling.

This is why mastery is about continuity across mathematical ideas—not learning a new formula in isolation.

Linear Programming Appears in Practical Systems

Manufacturing: allocate materials, labour and machine capacity.
Transportation: allocate flows through routes subject to capacity and cost.
Scheduling: distribute limited staff time among tasks.
Food production: choose combinations meeting nutrient or ingredient constraints.
Energy systems: allocate generation within capacity and demand requirements in suitable simplified models.
Supply chains: meet demand while controlling cost.

For more real-world detail, see Why Mathematics? | Supply Chains, Linear Programming and Delivery Planning.

Common Linear Programming Mistakes

Defining variables vaguely or inconsistently.
Using a profit objective when the given coefficients actually represent revenue.
Reversing ≤ and ≥ while translating words.
Forgetting non-negativity constraints.
Plotting the boundary lines correctly but shading the wrong half-plane.
Testing points that are not feasible.
Failing to include the intersection of two active constraints among the vertices.
Finding a numerical maximum but not explaining what the variables represent.
Rounding an answer to integers without checking the constraints.

Three Pathways for Building Linear Programming Mastery

Repair: Algebra and Inequality Language

If a student struggles with the topic, rebuild variables, linear equations and phrases such as “at most”, “at least”, “no more than” and “must exceed”. Practise identifying the quantity represented by each coefficient before trying full optimisation.

Stabilise: Feasibility and Corner Points

Use small two-variable models. Require the student to draw a feasible region, calculate every vertex, evaluate the objective and verify the selected point against the original resource restrictions. Do not accept a numerical answer without an interpretation sentence.

Extend: Real Decisions and Sensitivity

Introduce integer restrictions, infeasible systems, multiple optima, binding constraints, shadow prices and model sensitivity. Stronger learners can explore how computer algorithms solve higher-dimensional linear programmes.

A Weekly Linear Programming Routine

One translation: define variables and construct inequalities from a short story.
One graph: plot boundaries and shade the feasible region.
One intersection: solve two constraints simultaneously.
One objective comparison: evaluate all feasible vertices.
One reality check: verify resource use and interpret the answer.
One extension: change a constraint or objective and explain what happens.

Parent and Student Progress Checklist

I can define decision variables clearly.
I can distinguish objective functions from constraints.
I can translate “at most” and “at least” correctly.
I can graph linear boundaries.
I can identify the feasible region.
I can find intersection points.
I can calculate objective values at feasible vertices.
I can identify maximum or minimum solutions.
I can recognise binding and slack constraints.
I can explain infeasible and unbounded cases.
I know when integer restrictions matter.
I can explain the answer in the original real-world context.

Frequently Asked Questions

What is linear programming in simple terms?

It is a mathematical method for finding the best value of a linear objective while obeying linear constraints such as limits on resources, time, capacity or cost.

Why do we check corner points?

For a linear objective over a bounded polygonal feasible region, an optimum occurs at at least one vertex. Evaluating vertices therefore finds an optimum in standard two-variable graphical problems.

What is a feasible region?

It is the set of all choices satisfying every constraint simultaneously. A mathematically attractive point outside that region is not a valid solution.

Does linear programming require computer programming?

No. Small problems can be solved graphically and algebraically. Larger problems often benefit from specialised optimisation software.

Helpful Reading in the eduKateSG Ecosystem

The Core Aim

The core aim of linear programming mastery is not to make students draw more constraint lines or memorise a corner-point recipe. It is to make choices under limits mathematically defensible.

A strong learner can define the decision, model the available resources, identify which choices are feasible, locate the optimum and explain what that result means. That is mathematics doing one of its most practical jobs: helping us make better decisions when resources are limited.

The Core Aim of Mathematics Mastery | Linear Programming