CCW331

UNIT III

BUSINESS FORECASTING

Business Forecasting, Predictive Analytics

INTRODUCTION TO BUSINESS FORECASTING AND PREDICTIVE ANALYTICS

 Business forecasting refers to the tools and techniques used to predict
developments in business, such as sales, expenditures, and profits.
 The purpose of business forecasting is to develop better strategies based on these
informed predictions.
 Past data is collected and analyzed via quantitative or qualitative models so that
patterns can be identified and can direct demand planning, financial operations,
future production, and marketing operations.

The business forecasting process entails:

 Identify the problem, data point, or question that will be the basis of the
systematic investigation.
 Identify relevant, theoretical variables and determine the ideal manner for
collecting datasets.
 Make estimates about future business operations based on information collected
through investigation.
 Choose the model that best fits the dataset, variables, and estimates. The chosen
model conducts data analysis and a forecast is made.
 Note the deviations between actual performance and the forecast. Use this
information to refine the process of predicting and improve the accuracy offuture
forecasts.

UNIT-3- Page 1 of 33
CCW331

Business forescasting process:

 A problem or data point is chosen. This can be something like "will people buy a
high-end coffee maker?" or "what will our sales be in March next year?"
 Theoretical variables and an ideal data set are chosen. This is where the
forecaster identifies the relevant variables that need to be considered and decides
how to collect the data.
 Assumption time. To cut down the time and data needed to make a forecast, the
forecaster makes some explicit assumptions to simplify the process.
 A model is chosen. The forecaster picks the model that fits the dataset, selected
variables, and assumptions.
 Analysis. Using the model, the data is analyzed, and a forecast is made from the
analysis.
 Verification. The forecast is compared to what actually happens to identify problems,
tweak some variables, or, in the rare case of an accurate forecast, pat themselves on
the back.

UNIT-3- Page 2 of 33
CCW331

BusinessForecastingTechniques
Business forecasting and planning can be conducted by either quantitative modeling
methods or qualitative modeling methods:

Choosing a forecasting method depends on the following factors:

 Context
 Historical data: if it’s available and/or relevant
 Accuracy desired
 Timeperiod: short term vs. long term forecast
 Cost of forecasting vs. benefit
 Time available: do you need the forecast immediately?

Quantitative Techniques in Business Forecasting

 Quantitative forecasting is a long term business forecasting method
concerned only with measurable data such as statistics and historical data.
 Past performance is used to identify trends or rates of change. These types of
business forecasting are especially useful for long range forecasting in
business.
 Quantitative demand forecasting methods use mathematical models and
statistical techniques to make predictions based on numerical data, such as
sales, prices, income, population, and seasonality.
 These methods are often used when there is sufficient and reliable historical
data available, and when the demand follows a regular pattern or trend

UNIT-3- Page 3 of 33
CCW331

Quantitative models include:

Trend Analysis Method: also known as “Time Series Analysis,” this forecast method
uses past data to predict future events, excluding outliers and holding more recent data
in higher regard.
 This method is most effective when there is a large quantity of historical data showing
clear and stable trends. This is the most common and cost-effective method.
 Time series analysis requires a few years of data on either a specific product or
product line in order to get a clear sense of patterns.

Time series forecasting is a technique for the prediction of events through a sequence
of time. It predicts future events by analyzing the trends of the past, on the assumption
that future trends will hold similar to historical trends. It is used across many fields of
study in various applications including:
 Astronomy
 Business planning
 Control engineering
 Earthquake prediction
 Econometrics
 Mathematical finance
 Pattern recognition
 Resources allocation
 Signal processing
 Statistics
 Weather forecasting

UNIT-3- Page 4 of 33
CCW331

Time series models

Time series models are used to forecast events based on verified historical data. Common
types include ARIMA, smooth-based, and moving average. Not all models will yield the
same results for the same dataset, so it’s critical to determine which one works best based
on the individual time series.

When forecasting, it is important to understand your goal. To narrow down the specifics
of your predictive modeling problem, ask questions about:

 Volume of data available — more data is often more helpful, offering greater
opportunity for exploratory data analysis, model testing and tuning, and model
fidelity.
 Required time horizon of predictions — shorter time horizons are often easier to
predict — with higher confidence — than longer ones.
 Forecast update frequency — Forecasts might need to be updated frequently over
time or might need to be made once and remain static (updating forecasts as new
information becomes available often results in more accurate predictions).
 Forecast temporal frequency — Often forecasts can be made at lower or higher
frequencies, which allows harnessing downsampling and up-sampling of data (this in
turn can offer benefits while modeling).

Time series analysis vs. time series forecasting

While time series analysis is all about understanding the dataset; forecasting is all
about predicting it. Time series analysis comprises methods for analyzing time series
data in order to extract meaningful statistics and other characteristics of the data. Time
series forecasting is the use of a model to predict future values based on previously
observed values.

The three aspects of predictive modeling are:

 Sample data: the data that we collect that describes our problem with known
relationships between inputs and outputs.
 Learn a model: the algorithm that we use on the sample data to create a model that
we can later use over and over again.
 Making predictions: the use of our learned model on new data for which we don’t
know the output.

Types of forecasting methods

Model Use

Decompositional Deconstruction of time series

Smooth-based Removal of anomalies for clear patterns

UNIT-3- Page 5 of 33
CCW331

Moving-Average Tracking a single type of data

Exponential Smooth-based model + exponential window

Smoothing function

Examples of time series forecasting include: predicting consumer demand for a

particular product across seasons; the price of home heating fuel sources; hotel
occupancy rate; hospital inpatient treatment; fraud detection; stock prices

Decompositional models

Time series data can exhibit a variety of patterns, so it is often helpful to split a time series
into components, each representing an underlying pattern category. There are two main
types of decomposition: decomposition based on rates of change and decomposition
based on predictability.

Decomposition based on rates of change

 This is an important time series analysis technique, especially for seasonal

adjustment. It seeks to construct, from an observed time series, a number of
component series (that could be used to reconstruct the original by additions or
multiplications) where each of these has a certain characteristic or type of behavior.
 If data shows some seasonality (e.g. daily, weekly, quarterly, yearly) it may be useful
to decompose the original time series into the sum of three components:
Y(t) = S(t) + T(t) + R(t)
where S(t) is the seasonal component, T(t) is the trend-cycle component, and R(t) is
the remainder component.

There are several techniques to estimate such a decomposition. The most basic
one is called classical decomposition and consists in:

Estimating trend T(t) through a rolling mean

Computing S(t) as the average detrended series Y(t)-T(t) for each season (e.g. for each
month)

Computing the remainder series as R(t)=Y(t)-T(t)-S(t)

Time series can also be decomposed into:

 Tt, the trend component at time t, which reflects the long-term progression of the
series. A trend exists when there is a persistent increasing or decreasing direction in
the data.</span>
 Ct, the cyclical component at time t, which reflects repeated but non-periodic
fluctuations. The duration of these fluctuations is usually of at least two years.

UNIT-3- Page 6 of 33
CCW331

 St, the seasonal component at time t, reflecting seasonality (seasonal variation).

Seasonality occurs over a fixed and known time period (e.g., the quarter of the year,
the month, or day of the week).

It, the irregular component (“residuals” or "noise") at time t, which describes random,
irregular influences.

Additive vs. multiplicative decomposition

In an additive time series, the components add together to make the time series. In a
multiplicative time series, the components multiply together to make the time series.

Here is an example of a time series using an additive model:

An additive model is used when the variations around the trend do not vary with the
level of the time series.

Here is an example of a time series using a multiplicative model:

Smoothing-based models

In time series forecasting, data smoothing is a statistical technique that involves

removing outliers from a time series data set to make a pattern more visible.
Inherent in the collection of data taken over time is some form of random variation.
Smoothing data removes or reduces random variation and shows underlying trends and
cyclic components.

Moving-average model

In time series analysis, the moving-average model (MA model), also known as moving-
average process, is a common approach for modeling univariate time series. The
moving-average model specifies that the output variable depends linearly on the
current and various past values of a stochastic (imperfectly predictable) term.

Together with the autoregressive (AR) model (covered below), the moving-average
model is a special case and key component of the more general ARMA and ARIMA models
of time series, which have a more complicated stochastic structure.

Contrary to the AR model, the finite MA model is always stationary.

UNIT-3- Page 7 of 33
CCW331

Exponential Smoothing model

Exponential smoothing is a rule of thumb technique for smoothing time series data using
the exponential window function. Exponential smoothing is an easily learned and easily
applied procedure for making some determination based on prior assumptions by the
user, such as seasonality. Different types of exponential smoothing include single
exponential smoothing, double exponential smoothing, and triple exponential smoothing

Exponential smoothing is a broadly accurate forecasting method for short-term forecasts.

The technique assigns larger weights to more recent observations while assigning
exponentially decreasing weights as the observations get increasingly distant

Causal Methods

 These methods take into account other factors which affect your business. This
more advanced technique can work farther into the future than time series analysis.
It relies on having a large data set—a time series analysis plus market research.
 There are a variety of causal methods but regression analysis is the primary
method used to analyze cause-and-effect patterns.
 Regression analysis relates sales to other variables such as competition and the
economy.
 For example, a agricultural business may want to look at predicted weather patterns
over an upcoming season to determine whether their business will be affected
positively or negatively.
 Regression analysis can be performed using tools such as Excel or business owners
may want to invest in more sophisticated statistics software or even hire an expert to
perform this task.

In general, let us consider the forecast for a dependent variable Y using n independent
variables X1, X2, X3, … Xn. Then developing a forecasting logic requires establishing a
establishing as follows: Y= f(X1, X2, X3, … Xn)

Econometric Modeling:

This mathematical model makes use of several multiple-regression equations to test the
consistency of datasets over time and the significance of the relationship between
datasets, and to predict significant economic shifts and the potential effect of those shifts
on the company.

Indicator Approach:

This approach follows the relationship between certain indicators and uses the leading
indicator data in order to estimate the performance of the lagging indicators. Lagging
indicators are a type of KPI that measure business performance subsequently and
provide insight into the impact of business strategies on the results achieved. Other
quantitative techniques

UNIT-3- Page 8 of 33
CCW331

 Smoothing is a technique applied to time series to remove the fine-grained variation

between time steps.
 The hope of smoothing is to remove noise and better expose the signal of the underlying causal
processes. Moving averages are a simple and common type of smoothing used in time series
analysis and time series forecasting.
 Calculating a moving average involves creating a new series where the values are comprised of
the average of raw observations in theoriginal time series.
 A moving average requires that you specify a window size called the window width. This defines
thenumber of raw observations used to calculate the moving average value.
 The “moving” part in the moving average refers to the fact that the window defined by the
window width is slid along the time series to calculate the average valuesin thenew series.

Qualitative Techniques in Business Forecasting

Qualitative forecasting relies on industry experts or “market mavens” to make short-
term predictions. These techniques are especially useful in forecasting markets for
which there is insufficient historical data to make statistically relevant conclusions.
Human judgement is key and the challenge is to process this judgement in an unbiased,
logical way that results in quantitative estimates.

Qualitative models include:

1.Market Research:

 This is forecasting method that requires significant time, energy and resources.
 Information is collected via conversations with present and potential
customers about their needs for certain services or goods.
 Questionnaires, surveys and analysis of variables are all required to produce
accurate data. This information then needs to be analyzed while taking into account
limitations like the small data set.
 This method produces very accurate data for the short (less than three months) to
medium term (three months to two years) as it identifies changing customer
opinions
 Polls and surveys are conducted with a large number of prospective consumers
regarding a specific product or service in order to predict the margin by which
consumption will either decrease or increase.

UNIT-3- Page 9 of 33
CCW331

UNIT-3- Page 10 of 33
CCW331

2. Delphi Model:

 A panel of experts are polled on their opinions regarding specific topics. Their
predictions are compiled anonymously and a forecast is made.
 This structured approach brings together experts to answer a series of
questionnaires, with results from the previous questionnaire determining the
content for the next one. A facilitator is involved at each stage.
 For example, a freelance calligrapher wants feedback on branding. She probably
won’t have the resources to hire a panel of experts. However, she could post on a
LinkedIn group for creatives and ask solicit feedback about her logo, website etc.
 To get the best quality feedback, look for online forums that require a professional
credential to join and have effective administrators who keep the discussions
constructive and focused. Check the credentials and work experience of anyone
offering advice.
 In-House Expertise The staff member(s) with the most expertise on what’s being
forecasted take on the task. They use their in-depth knowledge to make
predictions.
 For example, the sales department is tasked with making a sales forecast thanks
to their intimate knowledge of their customers. In a small business, the owner will
likely need to step in.

The Delphi method generally involves the following stages:

 A panel of experts is assembled.

 Forecasting tasks/challenges are set and distributed to the experts.
 Experts return initial forecasts and justifications. These are compiled and
summarised in order to provide feedback.
 Feedback is provided to the experts, who now review their forecasts in light of the
feedback. This step may be iterated until a satisfactory level of consensus is reached.
 Final forecasts are constructed by aggregating the experts’ forecasts.
 Each stage of the Delphi method comes with its own challenges

UNIT-3- Page 11 of 33
CCW331

Setting the forecasting task in a Delphi

In a Delphi setting, it may be useful to conduct a preliminary round of information

gathering from the experts before setting the forecasting tasks. Alternatively, as experts
submit their initial forecasts and justifications, valuable information which is not shared
between all experts can be identified by the facilitator when compiling the feedback.

Feedback

Feedback to the experts should include summary statistics of the forecasts and outlines
of qualitative justifications. Numerical data summaries and graphical representations can
be used to summarise the experts’ forecasts.

Iteration

The process of the experts submitting forecasts, receiving feedback, and reviewing their
forecasts in light of the feedback, is repeated until a satisfactory level of consensus
between the experts is reached.

Final forecasts

The final forecasts are usually constructed by giving equal weight to all of the experts’
forecasts. However, the facilitator should keep in mind the possibility of extreme values
which can distort the final forecast

UNIT-3- Page 12 of 33
CCW331

Other techniques

3. The Sale Force Composite Method is a sale forecasting method wherein the sales
agents forecast the sales in their respective territories, which is then consolidated at
branch/region/area level, after which the aggregate of all these factors is consolidated
to develop an overall company sales forecast.

 The sales force composite method is the bottom-up approach where the sales
force gives their opinion on sales trend to the top management. Since, the
salesmen are the people, who are very close to the market, can give a more
accurate sales prediction on the basis of their experience with the direct
customers.

There are several advantages of sales force composite forecast method.

 The intimate knowledge and experience of the sales force in their respective
territories can be used efficiently.
 The responsibility to forecast sales rests on the shoulders of the sales agent and thus
could be held accountable if anything goes wrong.
 Since the sales agents forecast the sales by themselves, put more efforts to achieve
them.
 This method is more reliable because of a large population sample and moreover, it
can be readily broken down into product-wise, month-wise, area-wise forecast.

The sales force composite method is not free from the limitations too.

 Since the sales agents are not the experts in forecasting, they cannot employ the
sophisticated forecasting techniques properly and neither they have complete data to
have a fact-based forecasting.
 Also, the salesman often gets heavily influenced by the conditions existing in his
territory, due to which he either becomes more optimistic or more pessimistic about
the future sales.
 The sales agent might be well informed about all the conditions prevailing in his
territory, but may not be well equipped with the complete information about the
economic environment and the industry as a whole.

UNIT-3- Page 13 of 33
CCW331

 Sometimes, the sales agent intentionally gives fewer sales forecast, so that they can
fetch more incentives or bonus from the management on exceeding the sales targets.

4. Scenario writing forecast is essentially a backup sales forecasting method. Scenario

writing is used when a company’s long-term sales plan is difficult to predict due to
economic or environmental factors.

 The purpose of this method is to create various extremes that your sales data
may not be able to predict. This qualitative method for forecasting is used as a
long-term planning tool. For this forecasting method to be successful, sales and
company leaders need to have a strong subjective understanding of the sales cycle
and the company’s business model.
 Effective scenario writing needs to be built around scenarios that are created
to question uncertainties that may occur for your company. Scenario writing
forecasting isn’t complete until a company has developed a plan of action foreach
one of these possible scenarios.
 With this method, your goal is to project numerous likely outcomes based on
different, highly specific, assumption sets. So, you would draft several pictures of
what might unfold based on those individual sets.
 This provides you with the best case scenarios, worst case scenarios, and
everything in between.
 In order for scenario writing to be an effective forecasting technique, you need to
plan the scenarios in question around uncertainties that lie ahead for your
organization. Then, you need a clear plan of action that you would be able to
immediately implement should any of the projected scenarios happen.

With this technique there are eight steps you can follow in order to think
strategically about your scenarios and your planning process:

UNIT-3- Page 14 of 33
CCW331

 The focal issue within your assumption set.

 Key factors that are influenced by the focal issue and which influence the focal
issue.
 Any external forces from the focal issue.
 Critical uncertainties related to the external forces or critical uncertainties.
 Scenario logics, as they derive from: a) key factors b) external forces or c) critical
uncertainties.
 Scenarios
 Implications and options
 Early indicators, as they relate to the focal issue and the key factors

Eg:
If your focal issue is annual sales, your key internal factors might be sales calls or received
inquiries. The external factors that might impact this include local restrictions or
competitors.

Your critical uncertainties pertain to what issues you might face over the coming year.
This could include customers preferring online content rather than in-person lectures,
changes to local institutions that impact your business.

From there, you can develop scenarios that pertain to each of these potential situations,
in both good and bad directions. With that, you can develop the scenarios, what
implications it would have for the annual sales, what options your organization would
have if that event were to take place, and any early indicators.

5. Economic Indicator
Economic forecasting is the process of attempting to predict the future condition of the
economy using a combination of important and widely followed indicators.

Economic forecasting involves the building of statistical models with inputs of several
key variables, or indicators, typically in an attempt to come up with a future gross
domestic product (GDP) growth rate. Primary economic indicators include inflation,
interest rates, industrial production, consumer confidence, worker productivity, retail
sales, and unemployment rates.

1. Lagging Indicators
A lagging indicator is an observable economic variable that changes significantly
after a change has been observed in the real economy.

UNIT-3- Page 15 of 33
CCW331

Lagging indicators are used to inform which stage of the business cycle an economy
is in. They are also used to identify the overall trend of the economy and are used by
individuals, businesses, and government entities to make informed decisions.

Examples of lagging indicators are:

Gross Domestic Product (GDP) Growth Rate
Unemployment Rate
Consumer Price Index (CPI)
Central Bank Interest Rates
Corporate Earnings
Balance of Trade

The common characteristic among the lagging indicators is that the shift in them occurs
only after there has been a shock to the economy.

For example, during the 2008-2009 Global Financial Crisis, corporate earnings of
companies were not observed to have fallen until after the housing market bubble had
already popped.

2. Leading Indicators
A leading indicator is an observable economic variable that changes significantly
before a change has been observed in the real economy.

Leading indicators are used to predict when changes in the economic cycle may occur and
predict other significant shifts in the economy. As you can imagine, leading indicators are
critically important in economic forecasting since they are the main inputs in the
statistical models used to forecast economic conditions.

It should be noted that the data points are gathered from the past, and the past does not
necessarily inform future conditions. Therefore, leading indicators are not always
accurate, but they provide some insights and are widely used by individuals, businesses,
and government entities.

Examples of leading indicators are:

Yield Curve
Housing Starts
Retail Sales
Jobless Claims
Corporate Capital Expenditures
Purchasing Managers Index (PMI)
Consumer Confidence Index
Industrial Production
Worker Productivity

UNIT-3- Page 16 of 33
CCW331

6. Historical analogy
 Historical analogy method mainly forecasts the demand for a new product, it may be
accurate and cheap. It bases on forecasts and past data of any similar or relevant
existing products, then according to the product situations to develop a best fit
forecast.
 For examples, forecasting the demand of iPhone 6 phone cover, can base on the sales
of iPhone 6, forecasting the demand of iPhone 6 will base on the sales of iPhone 5; or
to forecasting the demand of an new type of camera film can base on the sales of the
company’s latest camera. The weakness of this is that it relies on analogy being
correct, and there is no guarantee that the new product demand will match.

PREDICTIVE ANALYTICS

 Predictive modelling is the method of making, testing and authenticating a model to

best predict the likelihood of a conclusion.
 Several modelling procedures from artificial intelligence, machine learning and
statistics are present in predictive analytics software solutions.

UNIT-3- Page 17 of 33
CCW331

 Models can utilise single or more classifiers to decide the probability of a set of data
related to another set.

The different models available for predictive analytics software enables the system to
develop new data information and predictive models. Each model has its own strengths
and weakness and is best suited for various types of problems.

 Predictive analytics represents the use of historical data to train machine learning
models to predict future outcomes based on computations made by statistical
algorithms.
 Organizations are turning to predictive analytics to solve business problems—
such as determining which products to market to specific customer segments and on
which platforms—and unearth new insights from their data.
 Mathematically speaking, predictive modeling involves approximating a mapping
function (f) from input variables (X) to output variables (Y).

 The most common use cases include optimizing marketing campaigns based on
historical customer data, detecting fraud by analyzing patterns of criminal behavior,
and forecasting inventory or setting prices.
 For example, airlines use predictive analytics to set ticket prices by forecasting seat
availability and demand.

Key points
 Predictive modelling is at the heart of business decision making
 Building decision models more than science is an art
 Creating an ideal decision model demands:
 Good understanding of functional business areas
 Knowledge of conventional and in-trend business practices and research
 Logical skillset

UNIT-3- Page 18 of 33
CCW331

 It is always recommended to start simple and keep on adding to the models as

required.

Applications
 Predictive models are used for all kinds of applications, including weather
forecasts, creating video games, translating voice to text, customer service, and
investment portfolio strategies. All of these applications use descriptive statistical
models of existing data to make predictions about future data.
 Predictive analytics is also useful for businesses to help them manage inventory,
develop marketing strategies, and forecast sales.
 It also helps businesses survive, especially those in highly competitive industries
such as health care and retail.
 Investors and financial professionals can draw on this technology to help craft
investment portfolios and reduce the potential for risk.

Uses of Predictive Analytics

Predictive analytics is a decision-making tool in a variety of industries.

 Forecasting
o Forecasting is essential in manufacturing because it ensures the optimal
utilization of resources in a supply chain. Critical spokes of the supply chain
wheel, whether it is inventory management or the shop floor, require accurate
forecasts for functioning.

o Predictive modeling is often used to clean and optimize the quality of data
used for such forecasts. Modeling ensures that more data can be ingested by the
system, including from customer-facing operations, to ensure a more accurate
forecast.

 Credit
Credit scoring makes extensive use of predictive analytics. When a consumer or
business applies for credit, data on the applicant's credit history and the credit record
of borrowers with similar characteristics are used to predict the risk that the applicant
might fail to perform on any credit extended.

 Underwriting
Data and predictive analytics play an important role in underwriting. Insurance
companies examine policy applicants to determine the likelihood of having to pay out
for a future claim based on the current risk pool of similar policyholders, as well as
past events that have resulted in payouts. Predictive models that consider
characteristics in comparison to data about past policyholders and claims are
routinely used by actuaries.

 Marketing
o Individuals who work in this field look at how consumers have reacted to the
overall economy when planning on a new campaign. They can use these shifts in

UNIT-3- Page 19 of 33
CCW331

demographics to determine if the current mix of products will entice consumers

to make a purchase.

o Active traders, meanwhile, look at a variety of metrics based on past events when
deciding whether to buy or sell a security. Moving averages, bands, and
breakpoints are based on historical data and are used to forecast future price
movements.

 Fraud Detection
Financial services can use predictive analytics to examine transactions, trends, and
patterns. If any of this activity appears irregular, an institution can investigate it for
fraudulent activity. This may be done by analyzing activity between bank accounts
or analyzing when certain transactions occur.

 Supply Chain
Supply chain analytics is used to predict and manage inventory levels and pricing
strategies. Supply chain predictive analytics use historical data and statistical models
to forecast future supply chain performance, demand, and potential disruptions. This
helps businesses proactively identify and address risks, optimize resources and
processes, and improve decision-making. These steps allow companies to forecast
what materials will be on hand at any given moment and whether there will be any
shortages.

 Human Resources
Human resources uses predictive analytics to improve various processes, such as
forecasting future workforce needs and skills requirements or analyzing employee
data to identify factors that contribute to high turnover rates. Predictive analytics can
also analyze an employee's performance, skills, and preferences to predict their
career progression and help with career development planning in addition to
forecasting diversity or inclusion initiatives.

 Predictive Analytics vs. Machine Learning

o A common misconception is that predictive analytics and machine learning are the
same things. Predictive analytics help us understand possible futureoccurrences
by analyzing the past. At its core, predictive analytics includes a series of statistical
techniques (including machine learning, predictive modeling, and data mining)
and uses statistics (both historical and current) to estimate, or predict, future
outcomes.

o Machine learning, on the other hand, is a subfield of computer science that, as per
the 1959 definition by Arthur Samuel (an American pioneer in the field of
computer gaming and artificial intelligence) means "the programming of a digital
computer to behave in a way which, if done by human beings or animals, would be
described as involving the process of learning."

Types of Predictive Analytics modeling

Predictive analytics models are classified into several types, including:

UNIT-3- Page 20 of 33
CCW331

 Regression models: These models use one or more input variables to predict a
continuous numeric value, such as sales revenue or customer lifetime value.
 Classification models: these are used to categorize data into one of two or more
groups based on input variables. Fraud detection, customer segmentation, and
spam filtering are a few examples.
 Time series models: These models are used to predict future values based on
historical trends and patterns in time-series data, such as stock prices, weather
patterns, or website traffic.
 Clustering models: These models are used to categorize data points based on
similarities in their characteristics or behaviors. Clustering models are oftenused
for customer segmentation and market basket analysis.
 Neural network models: These models are used to identify complex patterns in
data and are based on the structure and function of the human brain. Image and
speech recognition, natural language processing, and predictive maintenance are
all common applications for neural network models.
 Decision trees: These models are used to generate a visual representation of
possible outcomes based on various decision paths. They are frequently used to
assess risk, detect fraud, and analyze customer churn.
 Ensemble models: combine multiple predictive models to improve accuracy
while reducing the risk of overfitting. Random forests, gradient boosting, and
stacking models are some examples.
 These are a few predictive analytics models examples. There are numerous other
models, each with its own set of strengths and weaknesses, that can be used to
solve various types of problems.

The Advantages of Predictive Analytics Models

Businesses can benefit from predictive analytics models in a variety of ways, including:

 Improved Accuracy: Because predictive models incorporate more data and can
detect more complex relationships, they can provide more accurate predictions than
traditional statistical methods.
 Improved Decision-making: Predictive analytics models can assist businesses in
making more informed decisions based on data-driven insights rather than intuition
or guesswork.
 Increased Efficiency: By automating complex data analysis tasks, predictive
analytics models can help businesses save time and resources.
 Competitive Advantage: By identifying opportunities and making strategic
decisions ahead of their competitors, businesses that use predictive analytics models
can gain a competitive advantage.

The Limitations of Predictive Analytics Models

Although predictive models offer a wide range of benefits, they also a couple limitations
that must be considered. Here are some of the key limitations of predictive analytics
models.

UNIT-3- Page 21 of 33
CCW331

 Accurate predictions rely on accurate and complete data: Predictive analytics

models rely heavily on historical data. If the data used to build the model is inaccurate
or incomplete, it can lead to unreliable predictions. Additionally, if there are
significant changes in the data that the model has not accounted for, the model’s
accuracy may decrease over time.
 Flawed or inapt algorithms reduce the model’s accuracy: Secondly, predictive
analytics models are only as good as the algorithms used to create them. If the
algorithms are flawed or not suitable for the data being analyzed, the predictions may
not be accurate. It is essential to select the appropriate algorithm for the problem
being addressed and ensure that it is regularly updated and refined as new data
becomes available.
 Predictive models rarely account for outliers: Predictive analytics models may not
be able to account for unexpected events or outliers. As these events may not have
occurred in the inputted historical data, the predictive model will not be able to
predict outliers accurately. For this reason, it is pivitol to review the predictions
regularly and update the model when necessary to ensure its accuracy.

LOGIC AND DATA DRIVEN MODELS

1. The predictive analytics used in the BA process relate to the application of the
advanced statistical, software as well as the research methods so as to identify the
predictive set of variables. The model is built in order to identify the relations as
well as trends that are not generally seen with descriptive models.

2. The descriptive analysis can provide a relationship between the customer variables
but the precision of such estimation is not generally guaranteed.

3. The precision of establishing the sales and customer behavior relationship requires
some other kind of analysis tools. The predictive analysis on the other hand fares well
in this regard and satisfies the need for the exploring the data. The predictive analysis
establishes whether the particular relationship exist and develops the model in terms
of predicting the future events.

4. Predictive modeling means the developing models that can be used to forecast or
predict future events. Models can be developed either through logic or data.

o Logic driven models remain based on experience, knowledge and logical

relationships of variables and constants connected to the desired business
performance outcome situation.

o Data-driven Models refers to the models in which data is collected from

many sources to qualitatively establish model relationships. Logic driven
models is often used as a first step to establish relationships through data-driven
models. Data driven models include sampling and estimation, regression analysis,
correlation analysis, forecasting models and stimulation.

UNIT-3- Page 22 of 33
CCW331

Logic-Driven Models

A logic-driven model is one based on experience, knowledge, and logical relationships

of variables and constants connected to the desired business performance outcome
situation

Model building requires an understanding of business systems and the

relationships of variables and constants that seek to generate a desirable business
performance outcome. To help conceptualize the relationships inherent in a business
system, diagramming methods can be helpful.

For example, the cause-and-effect diagram is a visual aid diagram that permits a user to
hypothesize relationships between potential causes of an outcome

This diagram lists potential causes in terms of human, technology, policy, and process
resources in an effort to establish some basic relationships that impact business
performance.

The diagram is used by tracing contributing and relational factors from the desired
business performance goal back to possible causes, thus allowing the user to better
picture sources of potential causes that could affect the performance. This diagram is
sometimes referred to as a fishbone diagram because of its appearance.

Another useful diagram to conceptualize potential relationships with business

performance variables is called the influence diagram. Influence diagrams can be
useful to conceptualize the relationships of variables in the development of
models.

It maps the relationship of variables and a constant to the desired business performance
outcome of profit. From such a diagram, it is easy to convert the information into a
quantitative model with constants and variables that define profit in this situation:

Profit = Revenue − Cost, or

UNIT-3- Page 23 of 33
CCW331

Profit = (Unit Price × Quantity Sold) − [(Fixed Cost) + (Variable Cost × Quantity Sold)],
or

P = (UP × QS) − [FC + (VC × QS)]

It is necessary to be knowledgeable about the business systems being modeled in order

to capture the relevant business behavior. Cause-and-effect diagrams and influence
diagrams provide tools to conceptualize relationships, variables, and constants, but it
often takes many other methodologies to explore and develop predictive models.

 Logic driven models are created on the basis of inferences and postulations which
the sample space and existing conditions provide.
 Creating logical models require solid understanding of business functional areas,
logical skills to evaluate the propositions better and knowledge of business practices
and research.

To understand better, let us take an example of a customer who visits a restaurant around
six times in a year and spends around ₹5000 per visit. The restaurant gets around 40%
margin on per visit billing amount. The annual gross profit on that customer turns out to
be 5000 × 6 × 0.40 = 12000. 30% of the customers do not return each year, while 70% do
return to provide more business to the restaurant.

UNIT-3- Page 24 of 33
CCW331

Assuming the average lifetime of a customer (time for which a consumer remains a
customer) W 1/.3 = 3.33 years. So, the average gross profit for a typical customer turns
out to be 12000 × 3.33 = 39,960.

Armed with all the above details, we can logically arrive at a conclusion and can derive
the following model for the above problem statement:

Economic Value of each Customer (V) = (R × F × M)/D

Where,

R = Revenue generated per customer

F = Frequency of visits per year

M = Profit margin

D = Defection rate (Non-returning customers each year)

Data-Driven Models

Logic-driven modeling is often used as a first step to establish relationships through data-
driven models (using data collected from many sources to quantitatively establish model
relationships).

To avoid duplication of content and focus on conceptual material in the chapters, most
of the computational aspects and some computer usage content are relegated to the
appendixes

 The main aim of data-driven model concept is to find links between the state
system variables (input and output) without clear knowledge of the physical
attributes and behaviour of the system.
 The data driven predictive modelling derives the modelling method based on the set
of existing data and entails a predictive methodology to forecast the future
outcomes.
 It is data-driven only when there is no clear knowledge of the relationships among
variables/system, though there is lot of data. Here, you are simply predicting the
outcomes based on the data.
 The model is not based on hand-picked variables, but may contain unobserved,
hidden combination of variables.

UNIT-3- Page 25 of 33
CCW331

DATA MINING AND PREDICTIVE ANALYSIS MODELLING

Predictive analytics refers to the use of both new and historical data, statistical
algorithms, and machine learning techniques to forecast future activity, patterns, and
trends. The primary objective is to go beyond knowing what has happened to assess
better what will happen in the future.

On the other hand, data mining refers to the computational technique of discovering
patterns in huge data sets involving methods at the intersection of AI. The main

UNIT-3- Page 26 of 33
CCW331

objective of data mining is to extract useful information from the data warehouse and
transform it into a piece of useful information

Predictive' means to predict something, so predictive analytics is the analysis done to

predict the future event using the previous data. It is the process of extracting
information from existing sets of data to find useful information, trends and forecast
future events. Predictive analytics does not tell the exact thing that will happen in the
future. It predicts what might happen in the future.

Advantages of predictive analytics in business

 Predictive Analytics Increase production efficiency.

 It minimizes business risks.
 It helps in decision-making purposes in any business organization.
 It drives a competitive environment.

Strategies of Predictive Analytics

Based on the information gathered from predictive analytics, many companies have
increased their turnover, achieved goals, and increased revenues by applying these
strategies.

 Synchronizing supply with demand.

 Fraud prevention.
 Creating lasting inventories.
 Customer satisfaction.
 Setting prices appropriately in order to increase profit.

DataMining
Data mining refers to a process of analyzing data from different contexts and
summarizing it into useful information. The information gathered from data mining could
include customer patterns, purchase patterns, transaction times, customer demand, and
the relationship between the sold items.

UNIT-3- Page 27 of 33
CCW331

It is a powerful technology with great potential to assist companies in targeting themost

significant information in the data set they have gathered about the customer behaviour
and potential of the customers.

These are the given steps involved in the process of data mining

Business Understanings
 Data Selection
 Data Preparation
 Modelling
 Evaluation
 Deployment

Application of data mining

 Financial Analysis
 Biological Data Analysis
 Market Analysis
 Retail Industry
 Manufacturing Engineering
 Criminal investigation

Data mining model refers to a method that usually use to present the information and
various ways in which they can apply information to specific questions and problems. As
per the specialists, the data mining regression model is the most commonly used data
mining model.

In this process, a mining expert first analyzes the data sets and creates a formula that
defines them. Various Financial market analysts use this model to make predictions
related to prices and market trends.

UNIT-3- Page 28 of 33
CCW331

Another significant data mining model is based on the association rule. First, the data
mining analysts analyze the data sets to find which components usually appear together.
When they find the two components are paired simultaneously, it assumes that there are
some relation exits between them.

For instance, an electronic shop might find that consumers often purchase a marker and
pen simultaneously they purchase a book. A shop manager can use the detailed
information from the data mining model to increase sales by presenting all related
products at the same place.

Predictive Analytics Data Mining

Predictive analytics refers to the use of both data mining refers to the computational
new and historical data, statistical algorithms, technique of discovering patterns in huge
and machine learning techniques to forecast data sets involving methods at the
future activity, patterns, and trends intersection of AI.

It helps to make predictions based on future It helps to understand the gathered

events. information better.

Business analysts and other SMEs perform it. Statisticians and engineers perform it.

It applies business knowledge to find patterns It applies algorithms such asclassification

to get valid business predictions. and regression on gathered information
to find hidden patterns.

Types of Data mining models

 Predictive data mining models

 Descriptive data mining models


Predictive data mining models

 A predictive data mining model predicts the values of data using known results
gathered from the different data sets.
 Predictive modeling can not be classified as a separate discipline; it occurs in all
organizations or industries across all disciplines.
 The main objective of predictive data mining models is to predict the future based on
the past data, generally but not always on the statistical modeling.
 Predictive modeling is used in healthcare industries to identify high-risk patients with
congestive heart failures, high blood pressure, diabetes, infection, cancer, etc. It is also
used in the vehicle insurance company to assign the risk of accidents to the
policyholder

A predictive model of a data mining task comprises classification, regression, prediction,

and time series analysis. The predictive model of data mining is also called statistical
regression. It refers to a monitoring learning technique that includes an explication of

UNIT-3- Page 29 of 33
CCW331

the dependency of a few attribute's values upon the other attribute's value in the same
product and the growth of a model that can predict these attribute's values in previous
cases.

Classification:

In data mining, classification refers to a form of data analysis where a machine learning
model assigns a specific category to a new observation. It is based on what the model has
learned from the data sets. In other words, classification is the act of assigning objects to
many predefined categories.

One example of classification in the banking and financial services industry is identifying
whether transactions are fraudulent or not. In the same way, machine learning can also
be used to predict whether a loan application would be approved or not.

Regression:

Regression refers to a method that verifies the value of data for a function. Generally, it
is used for appropriate data.

A linear regression model in the context of machine learning or statistics is basically a

linear approach for modeling the relationships between the dependent variable known
as the result and your independent variable is known as features.

If your model has only one independent variable, it is called simple linear regression,
and else it is called multiple linear regression.

Types of regression

1. Linear Regression:

Linear regression is related to the search for the optimal line which fits the two
attributes so that with the help of one attribute, we can predict the other.

2. Multi-linear regression

Multi-linear regression includes two or more than two attributes, and the data are fit to
multi-dimensional space.

In data mining, prediction is used to identify data value based on the description of
another corresponding data value. The prediction in data mining is known as Numeric
Prediction.

Generally, regression analysis is used for prediction. For example, in credit card fraud
detection, data history for a particular person's credit card usage has to be analyzed. If
any abnormal pattern was detected, it should be reported as 'fraudulent action'.

UNIT-3- Page 30 of 33
CCW331

Time series analysis

Time series analysis refers to the data sets based on time. It serves as an independent
variable to predict the dependent variable in time.

Descriptive model
A descriptive model differentiates the patterns and relationships in data. A descriptive
model does not attempt to generalize to a statistical population or random process. A
predictive model attempts to generalize to a population or random process. Predictive
models should give prediction intervals and must be cross-validated; that is, they must
prove that they can be used to make predictions with data that was not used in
constructing the model.

Descriptive analytics focuses on the summarization and conversion of the data into useful
information for reporting and monitoring.

Clustering:

Clustering is grouping a set of objects so that objects in the same group called a cluster
are more similar than those in other groups clusters.

Association rules:

Association rules determine a causal relationship between huge sets of data objects. The
way the algorithm works is that you have. For example, a list of items you purchase at the
grocery store for the past six months data, and it calculates a percentage at which items
are purchased together. For example, what are the chances of you buying milk with
cereal?

Sequence:

Sequence refers to the discovery of useful patterns in the data is in relation to some
objective of how it is interesting.

Summarization:

Summarization holds a data set in more depth which is easy to understand form.

MACHINE LEARNING FOR PREDICTIVE ANALYTICS.

Predictive analytics (PA) and machine learning (ML) are powerful tools for
uncovering insights in large volumes of data. Organizations are using machine
learning to explore their large volumes of data and to automate processes. Machine
learning involves training algorithms, neural networks, or processing computers to
analyze data and output findings at scale. This output can include recommendations,
automated text, or flagged outliers

UNIT-3- Page 31 of 33
CCW331

Predictive analytics involves advanced statistics, including descriptive analytics,

statistical modeling and large volumes of data. Predictive analytics can include machine
learning to analyze data quickly and efficiently. Like machine learning, predictive
analytics doesn't replace the human element. Instead, PA supports data teams by
reducing errors and uncovering significant insights

Steps To Perform Predictive Analysis:

Some basic steps should be performed in order to perform predictive analysis.

 Define Problem Statement:

Define the project outcomes, the scope of the effort, objectives, identify the data sets
that are going to be used.
 Data Collection:
Data collection involves gathering the necessary details required for the analysis. It
involves the historical or past data from an authorized source over which predictive
analysis is to be performed.
 Data Cleaning:
Data Cleaning is the process in which we refine our data sets. In the process of data
cleaning, we remove un-necessary and erroneous data. It involves removing the
redundant data and duplicate data from our data sets.
 Data Analysis:
It involves the exploration of data. We explore the data and analyze it thoroughly in
order to identify some patterns or new outcomes from the data set. In this stage, we
discover useful information and conclude by identifying some patterns or trends.
 Build Predictive Model:
In this stage of predictive analysis, we use various algorithms to build predictive
models based on the patterns observed. It requires knowledge of python, R, Statistics
and MATLAB and so on. We also test our hypothesis using standard statistic models.
 Validation:
It is a very important step in predictive analysis. In this step, we check the efficiency
of our model by performing various tests. Here we provide sample input sets to check
the validity of our model. The model needs to be evaluated for its accuracy in this
stage.
 Deployment:
In deployment we make our model work in a real environment and it helps in
everyday discussion making and make it available to use.
 Model Monitoring:
Regularly monitor your models to check performance and ensure that we haveproper
results. It is seeing how model predictions are performing against actual data sets.

UNIT-3- Page 32 of 33
CCW331

UNIT-3- Page 33 of 33