Probability Theory is used to analyse data for decision making

Experiment is an operation that produces outcomes which can be observed

Outcome/event is the result of an experiment
Sample Space is the totality of all possible outcomes of an experiment
Events are said to be Mutually Exclusive when they cannot occur at the same time(occurrence of 1 implies n
another) when an experiment is performed.P(A) + P(A') = 1 and P(A and A')=0
Collectively Exhaustive list is a list of all possible events of an experiment

Classical Probability
probability is the ratio of n. of equalaly ikely outcomes favurable for an event to the total number of utcomes

Relative Frequency
we define probability of event A, occuring as the proportion of times A occurs, if we repeat the experiment sev
times under the same or similar conditions.

Subjective Probability
determined by a personal statement of how likely an outcome is.

Marginal Probability
ratio of number of possible outcomes favourable to the event A to th total number of possible outcomes
P(A) = number of possible outcomes favouring A / tota number of possible outcomes
the definition assumes that the elements of the sampe space have an equally likely chance of occuring.
Marginal probability in case of independent events is just the addition of the probabilities of all the events in w
simple event occurs.

Probability Rules

RULE 1 P(sample space S) = 1

RULE 2 0<=P(A)<=1
RULE 3 P(A or B) = P(A) + P(B) because P(A and B) = 0 when A and B are mutually exclusive evnts

Additional rule

In case of mutually exclusive events

P(A or B or C) = P(A) + P(B) + P( C)

In case events re not mutually exclusive
P(A or B) = P(a) + P(B) - P(A and B)

UNCONDITIONAL AND CONDITIONAL PROBABILITY

Independent and Dependent Events
Two events A and B are independent events if the occurrence of event A is in no way related to the occurrenc
occurrence of event B and vice versa
two events A and B are dependent events if the occurrence of event A is related to the occurrence event B

Multiplication rule: Indepnedent events Multiplication Rule : Dependent events

P(A and B) = P(A)*P(B) P(A and B) = P(A)*P(B/A)
P(B and A) = P(B)*P(A/B)
Conditional Probability: Inependent Events
if the probability of an event is subject to a restriction on the sample space, the probability is said to be conditi
conditional probability is the probability of the occurrence of an event say A, subject to the occurrence of a pre
P(A/B) = P(A)
Conditional Probability: dependent events
P(A/B) = P(A and B)/P(B)

Bayes' Theorem
P(Ai/B) = [ P(Ai)*P(B/Ai)] / [ P(A1)P(B/A1)+P(A2)P(B/A2)+….+P(Ak)P(B/Ak) ]
ame time(occurrence of 1 implies non occurrence of
=0

t to the total number of utcomes

s, if we repeat the experiment several

mber of possible outcomes

sible outcomes
ly likely chance of occuring.
e probabilities of all the events in which the

e mutually exclusive evnts

NAL PROBABILITY

in no way related to the occurrence or non-

sa
ated to the occurrence event B

Rule : Dependent events

he probability is said to be conditional.
subject to the occurrence of a previous event say B.
 Arithmetic Mean
X' = (X1 + X2 + … + Xn) / n
μ = Sum of all the data points in the population / size of popuation
X' = Sum of all the data points in the sample / Size of sample

Calculation for Discrete Series or Ungrouped Data

X' = ΣfX / Σf

Calculation for Continuous Series or Grouped Data

X' = Σfm / N
where,
m = mid-point of class = (Lower limit + Lower limit of next class) / 2
N = Σf

Median

ungrouped data grouped data

if there are odd number of items Median = {[(N+1)/2 - (F+1)]/fm} * W + Lm

media = (n+1/2) th item
if there are evn no. of items
mathematical property
the sum of the deviations of the items from
medians, ignoring signs, is the least as
compared to the sum of deviations of the
items from other value, ignoring signs
where
Lm = lower limit of the median class
fm = frequency of the median class
F = cumulative frequency upto the lower limit of the median
W = width of the class interval
N = Total frequency

Quartiles
those values which divide the total frequency into 4 equal parts.
Deciles
those values which divide the total frequency into 10 equal parts
Percentiles
those values which divide the total frquency into 100 equal parts
Any set of these partition values divides the area of the frequency
curve on histogram into equal parts

Mode
The mode is the value which occurs most frequently in a set of observations or the point of
maximum frequency and around which other items of the set cluster densely

Grouped Data
Mode = Lmo + (f - f1) / (2f - f1 - f2) * W
where,
Lmo = lower limit of the modal class which is the class having maximum frequency
f1, f2 = Frequencies of the classes preceding an succeeding the modal class respectively.
f = Frequency of the modal class
W = class interval

Bimodal distribution or multimodal distribution

if there are 2 or more values which have the maximum frequency, in such cases mode
should not be used as a measure of central tendency

Empirical Mode
Mode = 3Median - 2 Mean

MEASURES OF DIPERSION
The variability or dispersion of data of data is given by measures of dispersion.
When there is no dispersion, all the data points have identical values and the values of all the measures of central t

Range
range = value of highest data point - value of lowest data point

Deviation
average of the difference of the data points from a fixed value
the average deviation from zero = arithmetic mean
average deviation from arithmetic mean = 0

Mean Absolute Deviation

MAD = Σ|X - X'| / n

Standard deviation and Coefficient of Variation

Standard Deviation variance

steps square of standard deviation
1) calculation of deviations of the observations from the mean the smaller the value of
2) squaring each deviation
3) finding the mean of the squared deviation obtained step 2 A probability distribution is to future w
4) taking the positive square root of the mean found in step 3
σ = √ {[(X-μ)^2] / N}
where, Coefficient of Variation
X denotes each observation standard deviation / mean *100
μ is the arithmetic mean of the population
N is the number of observations
For Grouped Data Expected Standard deviation
σ = √ {[Σf(X-μ)^2] / Σf} σ = √ [Σpi(ki-k')^2]
where, f = frequency of the variable, μ = population mean where,
Step Deviation Method pi = Probbility associated with the oc
σ = i * √ { (Σfd^2/N) - (Σfd/N)^2 } ki = ith possible rate of return
where, k' = mean or expected rate of return
d = (m-A)/i n = number of possible outcomes
i = class interval
A = assumed mean
m = Mid-point of the class interval
Sample Standard Deviation
s = √ [Σ(X-X')^2] / [n-1] = √{ [ΣX^2/(n-1)] - [nX'^2 / (n-1)] }
where, s = Sample standard deviation
X' = Sample Mean
If each value is increased/ decreased by a constant quantity, Standard deviation remains same i.e. SD independent
If each value is multiplied/divided by a constant quantity, Standard Deviation will be similarly affected i.e. it is depen
Standard Deviation of a group
σ12 = √ { [N1σ1^2 + N2σ2^2 + N1d1^2 + N2d2^2] / [N1+N2]
where,
μi = mean of respective group
σi = standard deiation of i group
Ni = no. of observations in i group
d1 = μ1 -μ
d2 = μ2 - μ
μ = (N1μ1 + N2μ2) / (N1 + N2)
hmetic Mean
Mathematical Properties
1) Σ(X-X') = 0
2) Σ(X-X')^2 < Σ(X-A)^2 where A is any other point, different from X'
3) combined average
X'ab = (NaX'a + NbX'b) / ( NaNb)
where, X'ab = combined mean, X'a = mean of first group, Na = no. of iems in
first group, X'b = mean of secong group, Nb = no. of items in second group

weighted arithmetic mean

X'w = ΣwX / Σw

F+1)]/fm} * W + Lm

median class

cy upto the lower limit of the median class

Geometric Mean
nth root to the product of numbers to be averaged
G= The(X1*X2*X3…….Xn)^(1/n)
GM is used to find the average percent increase in sales,
production, population or other economic or business series
overtime.

Harmonic Mean
reciprocal of the arithmeic mean of the reciprocal of the given individual observations
HM = N/ [(1/X1) + (1/X2) + …..(1/Xn)]
Weighted Harmonic mean
WHM = ΣW / Σ(W/X)
 ES OF DIPERSION

lues of all the measures of central tendency converge.

Inter Quartile Range

last value in the third quartile - last value in the first quartile

Quartile Deviation
(Q3 - Q1) / 2
Q1 = Lq + [(N/4 - F)/fq] * W
Q3 = Lq + [(3N/4 - F)/fq] * W
where,
Lq = lower limit of the quartile class
N = total frequency
F = cumulative frequency upto quartile class, but not including quartile class
fq = frequency of the quartile class
W = width of the class interval

quare of standard deviation

he smaller the value of σ^2 the lesser the variability or greater the uniformity in the population.

probability distribution is to future what a frequency distribution is to past

Coefficient of Variation
tandard deviation / mean *100

xpected Standard deviation

= √ [Σpi(ki-k')^2]

i = Probbility associated with the occurrence of i th rate of return

i = ith possible rate of return
= mean or expected rate of return
= number of possible outcomes
remains same i.e. SD independent of change of origin
be similarly affected i.e. it is dependent on change of scale