DATA ANALYSIS VCE STUDENT RESOURCES

3/4 GENERAL MATHEMATICS 2023-2027 STUDY DESIGN

Types of Data Normal Distribution Causality

Categorical Nominal - no order Standard Deviation: how much members Common response (sunscreen &
Categorical Ordinal - have order from a group differ from the mean value fainting : hot weather temperature)
Numerical Discrete - how many? Normal Distribution: data distribution Confounding variables (two possible
Numerical Continuous - how much? with a bell shape 68-95-99.7% rule. explanatory variables)
For a normal distribution, approximately: Coincidence (divorce rate &
68% of the observations lie within margarine consumption)
Data Displays one standard deviation of the mean
95% of the observations lie within
There are many forms of ways to display
data. For categorical you can use...
two standard deviations of the mean Regression Line
99.7% of the observations lie within 3 The least squares line is the line that
frequency table
standard deviations of the mean minimises the sum of the squares of the
bar chart
Standard score = actual score - mean residuals. It is given by:
segmented bar cahrt
standard deviation y=a+bx
And for numerical data you can use...
frequency table b (slope)
histogram
dot plots
Population ‘r’ x standard deviation of y
standard deviation of x
stem and leaf plots Population: set of all eligible members of a (y-intercept)
a group which we intend to study mean value of y - ‘b’ x mean value of x
Sample: a subset of the population INTERPOLATION: Predicting within the
Log Scale range of values (reliable)
EXTRAPOLATION: Predicting outside
A log scale spreads out the countries
with small populations and 'pulls in' the
Bivariate Data range of values (unreliable)
RESIDUAL PLOT: residual value= actual
countries with huge populations. When investigating the association
b data value - predicted data value
If log10 x = b then 10 = x between two variables the explanatory
A logarithmic transformation involves variable (EV) is the variable we expect to
changing the scale on the horizontal axis
from x to log10(x), and replacing each of
explain or predict the value of the Transformations
response variable (RV).
the data values with its logarithm. Which Graph to use with Bivariate Data? y squared y squared
RV and EV are both categorical log. x x squared
Measures of centre segmented bar chart 1/x
RV and EV are both numerical
and spread scatterplot
log y
log x
1/y
log y
RV is numerical EV is categorical
Mean: ‘average’ : sum of all data values 1/x and 1/y x squared
parallel box plot/dot plot
divided by amount of values
back to back stem plot
Median: the middle value when placed in
ascending order Time Series data
Mode: most occuring number Scatterplot Patterns in this data include:
Range: the largest number subtract the
smallest number Things to note when interpreting: trends (increase/decrease)
IQR: spread of middle half of data values: Direction (positive/negative/none) cycles (recurrent movements)
quartile 3 - quartile 1 Outliers seasonality (related to calendar)
Form (linear or non-linear structural change (sudden change)
Strength (strong moderate or weak) possible outliers (stand out)

5 number summary irregular fluctuations (random)

SMOOTHING
The five number summary includes:
minimum (lowest number)
CC & CoD Can help to get rid of irregularities in
data/ seasonality.
quartile 1 (median of lower half) Correlation Coefficient: measures mean smoothing (calculate)
median strength of linear relationship (r), value is median smoothing (visual)
quartile 3 (median of upper half) between -1 (strong positive) and 1 SEASONALISING
maximum (largest number) (strong negative), 0 is no relation actual figure = deseasonalised figure x
A boxplot displays this. Coefficient of determination: degree to seasonal index
To find outliers... which one variable can be predicted Using a trend line fitted to a time series
UPPER FENCE: Q3 + 1.5 x IQR from another linearly related variable, plot to make predictions about future
LOWER FENCE: Q1 - 1.5 x IQR this is the squared value of r values is known as trend line forecasting.
FINANCE VCE STUDENT RESOURCES

3/4 GENERAL MATHEMATICS 2023-2027 STUDY DESIGN

Recurrence Relations Geometric Growth Finance Solver

SEQUENCE: a list of numbers written and decay A finance solver is a calculator
down in succession application that automates the
TERM: each of the numbers in a GEOMETRIC GROWTH: recurrence rule computing associated with analysing a
sequence multiplying by a factor greater than 1 reducing balance loan, an annutiy or an
A recurrence relation is a mathematical GEOMETRIC DECAY: recurrence rule annuity investment.
rule that we can use to generate a multiplying by a factor less than 1 N is the total number of payments.
sequence. It has two parts: Vn+1 = RVn I(%) is the annual interest rate.
starting value (first term in sequence) Another way is with an explicit rule: PV is the present value of the
n
rule (used to calculate next value) Vn = R x V0 loan/investment.
This can be written in symbols or words. Pmt is the amount paid at each
Words Symbols Compound interest is an example of payment.
Starting Value = 10 V0 = 10 geometric growth where any interest FV is the future value of the
Next term = Current + 5 Vn+1= Vn +5 that is earned is added to the principal loan/investment.
and then contributes to the earning of PpY is the number of payments per
interest in the next time period. year.
Linear Growth/Decay For this R= 1+ r/100 If you receive money, or someone owes
you money, we treat this as a positive
LINEAR GROWTH is when a recurrence Reducing Balance Depreciation is an (+ve) if you pay out money or you owe
rule involves adding a fixed amount. example of geometric decay where the someone money, we treat this as a
Vn+1 =Vn+ D value of an item is reduced by the same negative (−ve).
where ‘D’ is the common difference percentage every year. COMPOUND INTEREST INVESTMENT
Another way is with an explicit rule: For this R= 1- r/100 PV is negative
V n= V0 + nD Pmt is negative
Simple interest investments are an FV is positive
example of linear growth. They use the
same rule as linear growth except D is
Linear & Geometric REDUCING BALANCE LOAN
PV is positive
r/100 x principle (r = interest rate) a recurrence relation of the form below Pmt is negative
Principle: the initial amount borrowed, can be used to model situations that FV is negative, positive or zero
lent or invested involve both geometric and linear. ANNUITY
Interest: the amount of money V0= Starting Value, Vn+1= R x Vn ± D PV is negative
paid/earned for borrowing/lending Examples of this include: Pmt is positive
money over a period of time Reducing Balance Loans: borrower FV is positive or zero
makes regular payments to reduce the
LINEAR DECAY is when a recurrence rule amount owed, rather than waiting until
involves subtracting a fixed amount. the end of the loan to repay the balance Interest-only loans
Vn+1 =Vn - D Annuity: an investment where
Another way is with an explicit rule: compound interest is earned & money is Interest only loan: a loan in which only
Vn = V0- nD withdrawn from the investment in the the interest is paid.
Depreciation is a type of linear decay. form of regular payments The balance of the loan remains the
It’s the reduction in value of an item. same for the duration of the loan as
Flat rate: value of an item is reduced there is no money being paid off the
by the same amount each year Amortisation Tables principle
Unit-cost: reduced value based on Modelling interest only loans:
Amortisation tables charts the
how much the item is used V n+1 = R x Vn- D
repayments of a reducing balance loan
Scrap Value: value at which the item is or annuity on a step-by-step basis or the
no longer of use to the business payment of a compound interest
investment with additional payments.
Perpetuities
An example is shown below for a loan of Perpetuity: an investment where an
Interest over time $1000 with interest at 15% per annum equal amount is paid out on a regular
and monthly repayments of $250 basis forever
A nominal interest rate is converted by
Payment Interest Principle Balance Perpetuities have same relationship
dividing by the compounding period.
Reduction to annuities as interest-only loans
The ‘growth multiplier’, ‘R’ is now...
0.00 0.00 0.00 1000 have to reducing balance loans.
R= 1+ r/100xp (p is the no. of
250.00. 12.50. 237.50. 762.50 the value of the investment never
compounding periods each year)
250.00. 9.53. 240.47. 522.03 decreases because you are only
effective interest rate: the total interest
250.00. 6.53. 243.37. 278.56 being paid interest
in one year/principle x 100%
MATRICES VCE STUDENT RESOURCES

3/4 GENERAL MATHEMATICS 2023-2027 STUDY DESIGN

Matrix Basics Inverse/Determinant Transition Recursion

A matrix is a rectangular array or table of Inverse Matrix: A matrix that when Recurrence relation (similar to finance)
numbers or symbols, arranged in rows multiplied by the original matrix gives an must have a starting point, this is the
and columns. identity matrix. initial state matrix as well as the
ORDER OF A MATRIX Determinant: associated with square transition matrix as shown below.
number of rows × number of columns matrices, it is used to decide if a matrix
S 0= initial value, S n+1 = T S n
has an inverse. If the determinant is 0,
Using a general rule this can be written:
the matrix has no inverse (it is singular)
Types of Matrices HOW TO FIND THE DETERMINANT
n
S n= T S0
Steady / Equilibrium State: this occurs in
Row matrix: contains one a b
7324 det (A) axd-bxc the long term when their is no longer a
row of numbers c d
change when the transition matrix is
Column matrix: contains 7
applied. This means the values remain
one column of numbers 3
Transposed matrix: rows 1 2 1 3 Matrices with 0 & 1s the same from that point forward. To
solve for this we use large values for ‘n’
& columns are switched 3 4 2 4
1 2 Binary Matrices: a matrix which only
Square matrix: columns &
consists of zeros or ones More complex problems use this formula
rows have equal number 3 4
Permutation Matrices: a square binary S 0= initial value, S n+1 = T S n +B
Diagonal matrix: all 1 0 0
0 4 0 matrix, with only a singular one in each where ‘B’ is a column matrix
elements off the leading
0 0 6 row and column
matrix are zero
Communication Matrices: a square
Identity matrix: diagonal
matrix but the numbers
1 0 0
0 1 0
binary matrix where the ones represent Leslie Matrices
0 0 1 direct (one-step) communication links
on diagonal are all one. Leslie matrices are used to construct
NOTE: redundant communication links
Symmetric matrix: matrix 1 2 6 discrete models of population growth. In
are those in which the sender and
that is unchanged by a 2 4 5 particular, they are used to model
receiver are the same person.
transposition 6 5 7
changes in the sizes of different age
Triangular matrix: a groups within a population.
1 2 6
matrix where all elements
above/below are 0.
0 4 5
0 0 7
Dominance Matrices Note: Only the females of the species
are counted in the population, as they
One step dominances: The first are the ones who give birth to the new
dominance matrix, records the number members of the population.
Matrix Arithmetic of one-step dominances. This can
calculate a 1-step dominance score, by Separated into age groups and takes into
ADDITION & SUBTRACTION
summing each of the rows of the matrix. account two factors:
Both matrices have to have the same
Two step dominances: occurs when a birth rate (avg number of female
order. Then add/ subtract the
player beats another player who has offspring e.g 1.7)
corresponsing elements.
beaten someone else. Calculate by the survival rate (has to be a number
1 2 5 6 6 8
square the one-step dominance matrix. between 0 and 1)
3 4 3 2 6 6
5 6 1 2 4 4
An example is shown below
3 2 3 4
SCALAR MULTIPLICATION
0 -2
Transition Matrices There are three age groups. Age group 2s
birthrate is 2.3 and group 3s is 0.4. 60%
When you mutliply a matrix by a number, A transition matrix is a matrix that survive from group 1 to 2 whereas only
every element is multiplied by it. describes the way in which transitions 30% survive from group 2 to 3.
5 6 15 18 are made between two states. The
3
3 2 9 6 numbers are written in a proportion. 1 2 3
MULTIPLYING TWO MATRICES Below is an example of how a diagram 1 0 2.3 0.4
Matrix multiplication requires the no. of can be represented in a matrix.
columns in the 1st matrix to equal the 2 0.6 0 1
no. of rows in the 2nd matrix. The 20% 3 0 0.3 0
product matrix will have the same no. of 80% A B 90%
rows as the 1st matrix and the same no. 10%
of columns as the second matrix. A B 2.3 0.4
1 2 3
1 0 2 1x2+0x3 2 A 0.8 0.1 0.6 0.3
2 3 3 2x2+3x3 13 B 0.2 0.9
NETWORKS VCE STUDENT RESOURCES

3/4 GENERAL MATHEMATICS 2023-2027 STUDY DESIGN

Network Basics Adjacency Matrix Trees

A graph consists of vertices (dots) A square matrix showing the number of A tree is a connected graph that has no
joined by edges (lines). edges joining each pair of vertices in a loops, multiple edges or cycles.
number of edges attached to a vertex graph. D A B C D The number of edges is always one less
B
is called the degree of the vertex. A 1 1 0 0 than the number of vertices.
A loop connects a vertex to itself. A B 1 0 1 1 A spanning tree is a tree that connects
Loops contribute two degrees to a C 0 1 0 2 all of the vertices of a graph.
C
vertex. D 0 1 2 0 PRIMS ALGORITHM
An algorithm to help determine the

Types of Graphs Travelling

minimum spanning trees in a graph.
1. look for the smallest weight on an
Simple graphs - no loops or multiple / edge and include this
Walk- starts at one vertex, follows any
duplicate edges 2. add the next smallest weight of an
route and finishes at another vertex
edge to your tree
Trail- a walk with no repeated edges
Isolated Vertex- a vertex that is not 3. continue until all of the vertices are
Path- a trail with no repeated vertices
connected to another vertex by an edge connected (do not include an edge if
Circuit- trail with same start & end vertex
it forms a loop or cycle on the tree)
Cycle- path with same start & end vertex
Eulerian trails- follows every edge of a
Degenerate graphs- all vertices are
isolated (no edges in graph)
graph and is connected and has exactly Euler’s Formula
0 or 2 vertices that have an odd degree
Connected Graphs- no isolated vertices
Eulerian circuits- eulerian with same For any planar graph:
(every vertex connected)
start & end vertex, is connected and has v+f=e+2
all vertices of an even degree ‘v’ are the vertices, ‘f’ are the faces and
Bridge- an edge that makes the graph
Hamilton path- visits every vertex ‘e’ are the edges
stay connected
Hamilton cycle- Hamilton path that
starts and ends at the same vertex
Complete Graph- their is an edge Hungarian Algorithm
between every set of vertices
Flow problems Algorithm to figure out the most
efficient way to allocate vertex to
Equivalent Graph- show same
information, the location of verticies and A cut is an imaginary line across a another vertex in bipartite graphs.
edges are different directed graph that completely 1.Subtract the lowest value in each row,
separates the source (start of the from every value in that row.
flow) from the sink (end of flow). 2. If a column does not contain a zero,
The cut capacity is the sum of the subtract lowest value in that column
Planar Graph- a graph that can be drawn capacities of the edges that are cut. from every value in that column.
with no intercepting edges Only edges that flow from the source 3. Draw minimum lines over the zeros.
side to the sink side of the cut are Add the smallest uncovered value to any
included in a cut capacity. value that is covered by two lines.
The minimum cut capacity possible Subtract the smallest uncovered value

Dijksta’s algorithm for a graph equals the maximum flow

through the graph.
from all the uncovered values.
NOTE: If at any any point minimum
Algorithm to help you find the shortest number of lines required to cover zeros
path in a weighted graph. is equal to number of allocations to be
1. assign the starting vertex value of 0 Activity Networks made you can STOP, you are done and
and circle can now draw your bipartite graph and
Immediate predesessor is an activity
2. write the weight of all edges make your allocation.
that must be completed before
connected to 0 on the next vertex
another one can start.
3. circle the smallest value
4. with all vertices connected to the
Presedence tables record the
activities, EST, LST, duration and IP.
Scheduling Problems
circled one, add the next edges
Activity networks have edges that Projects that involve multiple activities.
weight and write on the next vertex
represent activities. Vertices aren’t A critical path is the longest or equal
5. circle the smallest value
labelled (except start and finish) longest path in an activity network. Float
6. continue step 4 & 5 until the end
Dummy activities required if two time is the latest starting time (LST)
vertex has been circled
activities share some, but not all, of subtract the earliest starting time (EST).
NOTE: Once vertex has been circled, it
their immediate predecessors The process for determining the critical
can’t be changed. Replace larger values
(dotted line) - weight is always zero path is called critical path analysis.
with the smaller values if un-circled.