MODULE-1

Introduction to Marketing Analytics and Data

Mining

Introduction to Marketing Analytics

Marketing Analytics involves the use of data and technology to measure, manage, and
analyze marketing performance. The goal is to maximize the effectiveness of marketing
efforts and optimize return on investment (ROI). By leveraging various analytical tools
and techniques, businesses can gain insights into consumer behavior, measure the
impact of marketing campaigns, and make data-driven decisions.

Need of Marketing Analytics

1. Data-Driven Decision Making: With vast amounts of data available, businesses

need to use analytics to make informed decisions rather than relying on intuition
or guesswork.
2. Customer Insights: Analytics helps in understanding customer preferences,
behaviors, and trends, which is crucial for targeting and personalization.
3. Measuring Effectiveness: It enables the measurement of the performance of
marketing campaigns, helping to identify what works and what doesn't.
4. Optimizing Spend: By analyzing ROI and other key metrics, companies can
allocate their marketing budgets more effectively.
5. Competitive Advantage: Companies that leverage marketing analytics can stay
ahead of competitors by quickly adapting to market changes and consumer
needs.

Benefits of Marketing Analytics

1. Improved Targeting: By analyzing customer data, businesses can segment their

audience and tailor marketing messages to specific groups, increasing the
effectiveness of campaigns.
2. Enhanced Customer Experience: Understanding customer preferences and
behaviors allows for more personalized and relevant interactions, improving
customer satisfaction and loyalty.
3. Increased ROI: Analytics helps in optimizing marketing spend by focusing on
high-performing channels and tactics.
 4. Better Decision Making: Data-driven insights lead to more strategic decisions,
reducing risks and improving outcomes.
5. Performance Tracking: Continuous monitoring and analysis of marketing
activities enable businesses to track progress and make necessary adjustments
in real time.

Data Mining - Definition

Data mining is the process of discovering patterns and knowledge from large amounts
of data. The data sources can include databases, data warehouses, the internet, and
other data repositories. The goal of data mining is to extract useful information that can
be used for decision making.

Classes of Data Mining Methods

1. Grouping Methods: These methods involve segmenting data into clusters or

groups based on similarities. Common techniques include:
● Clustering: Identifies distinct groups within the data without pre-defined labels.
● Association Rules: Discovers interesting relationships between variables in large
databases, often used in market basket analysis.

2. Predictive Modeling Methods: These methods predict future outcomes based

on historical data. Techniques include:
● Regression Analysis: Predicts a continuous outcome based on one or more
predictor variables.
● Classification: Assigns items to predefined categories or classes, such as spam
detection or credit scoring.
 3. Linking Methods to Marketing Applications: Data mining methods can be
directly applied to various marketing applications:
● Customer Segmentation: Using clustering to segment customers based on
purchasing behavior.
● Churn Prediction: Using classification to predict which customers are likely to
leave.
● Cross-Selling and Up-Selling: Using association rules to identify product bundles
or suggest complementary products.

Process Model for Data Mining - CRISP-DM

The Cross-Industry Standard Process for Data Mining (CRISP-DM) is a widely used
methodology for data mining projects. It consists of six phases:

1. Business Understanding: Determining business objectives, assessing the

situation, defining data mining goals, and producing a project plan.
2. Data Understanding: Collecting initial data, describing data, exploring data, and
verifying data quality.
3. Data Preparation: Selecting data, cleaning data, constructing data, integrating
data, and formatting data.
4. Modeling: Selecting modeling techniques, generating test design, building
models, and assessing models.
5. Evaluation: Evaluating results, reviewing processes, and determining next steps.
6. Deployment: Planning deployment, monitoring and maintaining models, and
producing final reports.

CRISP-DM provides a structured approach to ensure that data mining projects are
successful and that the results are actionable and beneficial for the business.

---------------------------------------------------------------------------------------------------------------------
 MODULE-2
Introduction to R

R: Overview

R is a programming language and free software environment used for statistical

computing and graphics. It is widely used among statisticians and data miners for data
analysis and developing statistical software. R is highly extensible and provides a wide
variety of statistical (linear and nonlinear modeling, classical statistical tests, time-series
analysis, classification, clustering) and graphical techniques.

Data Types and Structures in R

Data Types:

● Numeric: Represents numbers, including integers and doubles (e.g., 42, 3.14).
● Integer: A subset of numeric that represents whole numbers (e.g., 42L).
● Character: Represents text strings (e.g., "Hello, World!").
● Logical: Represents boolean values (TRUE or FALSE).
● Factor: Represents categorical data (e.g., factor(c("Male", "Female"))).
● Complex: Represents complex numbers (e.g., 1+2i).

Data Structures:

● Vector: A sequence of data elements of the same basic type. Example: c(1, 2, 3,
4).
● Matrix: A two-dimensional array where each element has the same type.
Example: matrix(1:9, nrow=3).
● Array: A multidimensional array. Example: array(1:8, dim=c(2,2,2)).
● List: A collection of elements possibly of different types. Example: list(1, "a",
TRUE, 1+2i).
● Data Frame: A table or two-dimensional array where each column can contain
different types of data. Example: data.frame(id=1:4, name=c("A", "B", "C", "D")).

Data Coercion

Data Coercion refers to the conversion of data from one type to another. In R, this can
happen implicitly or explicitly.

● Implicit Coercion: When performing operations, R will automatically coerce data

to a compatible type. For example, combining numeric and character vectors will
coerce the numeric values to characters.
● Explicit Coercion: Using functions to deliberately change the type of data
 Data Preparation: Merging, Sorting, Splitting, Aggregating

Merging: Combining data from two or more data sets based on a common key.

● Inner Join: Merges data that have matching keys in both datasets.
● Outer Join: Merges all data, filling with NA where there are no matches.
● Left Join: All data from the left dataset, and matching data from the right dataset.
● Right Join: All data from the right dataset, and matching data from the left
dataset.

Sorting: Arranging data in a particular order.

● Use sort() for vectors and order() for data frames.

● Example: sort(c(3, 1, 4)) results in 1, 3, 4.

Splitting: Dividing data into subsets based on some criteria.

● Use split() function to split data into groups.

● Example: split(iris, iris$Species) splits the iris dataset by species.

Aggregating: Summarizing data by grouping and applying a function.

● Use aggregate() function to perform summary statistics.

● Example: aggregate(mpg ~ cyl, data=mtcars, FUN=mean) computes the mean
mpg for each number of cylinders.

Introduction to R Libraries

R Libraries are collections of functions and datasets developed by the community. They
extend R's capabilities.

Installing Libraries:

● Use the install.packages("package_name") function.

● Example: install.packages("ggplot2") installs the ggplot2 package.

Invoking Libraries:

● Use the library(package_name) function to load the package.

● Example: library(ggplot2) makes ggplot2 functions available in your session.
 Introduction to R Graphs

R provides extensive graphical capabilities for data visualization.

1. Basic R Charts:
● Scatter Plot: plot(x, y) for simple scatter plots.
● Line Plot: plot(x, y, type="l") for line charts.
● Bar Plot: barplot(height) for bar charts.
● Histogram: hist(x) for histograms.
● Box Plot: boxplot(x) for box plots.

2. Different Types of Charts:

● Scatter Plot: Used for visualizing the relationship between two continuous
variables.
● Line Plot: Used for trends over time.
● Bar Plot: Used for categorical data comparisons.
● Histogram: Used for the distribution of a single continuous variable.
● Box Plot: Used for comparing distributions across groups.

---------------------------------------------------------------------------------------------------------------------
 MODULE-3

Descriptive Analytics and Application of Analytics in

Marketing

1. Exploratory Data Analysis (EDA)

Exploratory Data Analysis is the process of analyzing data sets to summarize their main
characteristics, often using visual methods. It helps to uncover patterns, spot anomalies,
test hypotheses, and check assumptions with the help of summary statistics and
graphical representations.

Summary Table: Provides a quick overview of the data, including measures such as
mean, median, standard deviation, minimum, and maximum values for each variable.

Charts and Graphs:

● Histograms: Show the distribution of a single numerical variable.

● Box Plots: Display the distribution of a numerical variable and identify outliers.
● Scatter Plots: Illustrate relationships between two numerical variables.
● Bar Charts: Represent categorical data with rectangular bars.
● Heatmaps: Show data magnitude as color in a matrix format.

Slicing and Dicing refers to the ability to analyze data from different viewpoints, often
using techniques such as pivot tables. This method helps in understanding specific
segments of data and identifying trends and patterns.

Inferential Statistics

Inferential Statistics allows us to make predictions or inferences about a population

based on a sample of data.

1. T-Test: Used to determine if there is a significant difference between the means

of two groups. For example, comparing the average sales before and after a
marketing campaign.
2. ANOVA (Analysis of Variance): Used to compare the means of three or more
samples. For example, evaluating the effectiveness of different marketing
strategies on sales.
 3. Chi-Square Test: Used to determine if there is a significant association between
two categorical variables. For example, examining the relationship between
gender and product preference.

Correlation measures the strength and direction of a linear relationship between two
variables. The correlation coefficient ranges from -1 to 1, where:

● 1 indicates a perfect positive relationship,

● -1 indicates a perfect negative relationship,
● 0 indicates no linear relationship.

Association Rules - Market Basket Analysis

Market Basket Analysis is a technique used to uncover associations between items in

large datasets, often used in retail to understand customer purchasing behavior.

● Association Rules: These rules identify items that frequently co-occur in

transactions. For example, if customers frequently buy bread and butter together,
the rule {bread} -> {butter} may be established.
● Support: The proportion of transactions that include both items.
● Confidence: The likelihood that a transaction containing one item also contains
the other.
● Lift: The ratio of observed support to that expected if the two items were
independent.

RFM Analysis

RFM Analysis is a customer segmentation technique based on three key metrics:

● Recency (R): How recently a customer made a purchase.

● Frequency (F): How often a customer makes a purchase.
● Monetary (M): How much money a customer spends on purchases.

Customers are scored on each of these dimensions, and these scores are used to
identify the most valuable customers and tailor marketing efforts accordingly.
 Customer Segmentation using K-Means Cluster Analysis

Customer Segmentation involves dividing a customer base into distinct groups that
share similar characteristics.

● K-Means Clustering: A method to partition data into K distinct clusters based on

similarity. The algorithm minimizes the variance within each cluster while
maximizing the variance between clusters.
● Customers are grouped based on selected features (e.g., age, spending
behavior).
● Each customer is assigned to the cluster with the nearest mean.

Key Driver Analysis using Regression Model

Key Driver Analysis identifies the factors that most significantly impact a given outcome,
often using regression models.

● Regression Analysis: A statistical technique for estimating the relationships

among variables.
● Linear Regression: Models the relationship between a dependent variable and
one or more independent variables by fitting a linear equation to the observed
data.
● Multiple Regression: An extension of linear regression that models the
relationship between a dependent variable and several independent variables.

In marketing, regression models can identify which factors (e.g., advertising spend,
pricing strategy) most strongly drive sales or customer satisfaction.

------------------------------------------------------------------------
 MODULE-4
Prediction and Classification Modelling using R

Introduction to Prediction and Classification Modelling

Prediction Modeling:

Prediction modeling involves using statistical techniques to create a model that can
predict future outcomes based on historical data. The goal is to develop a predictive
model that provides a quantitative output, such as predicting sales figures or stock
prices.

Classification Modeling:

Classification modeling, on the other hand, is used to categorize data into predefined
classes or groups. For instance, it can be used to classify whether a customer will churn
(leave a service) or not based on their behavior and characteristics.

Data Splitting for Training and Testing Purpose

Data splitting is a crucial step in building predictive and classification models. The
dataset is typically divided into two parts: a training set and a testing set.

● Training Set: This subset of data is used to train the model. The model learns
the underlying patterns and relationships in the training data.
● Testing Set: This subset is used to evaluate the performance of the model. The
model makes predictions on this data, and the accuracy and other performance
metrics are calculated.

The purpose of splitting the data is to ensure that the model can generalize well to
unseen data and is not overfitted to the training data.

Prediction Modeling

1. Predicting Sales Using Moving Average Model:

The Moving Average Model is a simple time series forecasting method. It involves
averaging a fixed number of past observations to make future predictions.

Simple Moving Average (SMA): This involves taking the average of a fixed number of
previous periods to smooth out short-term fluctuations and highlight longer-term trends
 2. Predicting Sales Using Regression Models:

Regression models establish the relationship between a dependent variable (e.g.,

sales) and one or more independent variables (e.g., advertising spend, price).

Simple Regression Model:

Multiple Regression Model:

Classification Modeling

1. Customer Churn Using Binary Logistic Regression:

Binary Logistic Regression is used when the dependent variable is binary (e.g., churn or
not churn).

.
 2. Customer Churn Using Decision Tree:

Decision trees classify data by splitting it into branches based on the values of the input
features.

● A decision tree consists of nodes (decision points) and leaves (final

classifications). Each node represents a feature, and each branch represents a
decision rule.
● The tree grows by recursively splitting the data at each node based on the
feature that results in the greatest information gain (or another splitting criterion)
until the stopping criteria are met (e.g., maximum depth or minimum samples per
leaf).
● For customer churn, the decision tree will split the data based on features like
customer tenure, usage patterns, and other relevant factors to classify whether a
customer is likely to churn or not.

Decision trees are intuitive and easy to visualize but can be prone to overfitting.
Techniques like pruning (removing parts of the tree that provide little power) are used to
improve their performance.

Summary

In summary, prediction models (such as Moving Average and Regression models) focus
on forecasting numerical outcomes, while classification models (such as Binary Logistic
Regression and Decision Trees) focus on categorizing data into discrete classes. Both
types of models require proper data splitting to ensure that they generalize well to new,
unseen data.