Department of Computer Science and Engineering

(Data Science)
Haldia Institute of Technology

NIELIT, Kolkata
Work-Based Learning
(WBL)
Name: Baridhara Ghosh
Roll no.: 10330521036
Supervisor: Mr. Bhaskar Banerjee (Scientist-D)
 Acknowledgement
I sincerely thank NIELIT, Kolkata and my supervisor, Mr. Bhaskar
Banerjee, for their guidance and support during my internship. I am
also grateful to my faculty at Haldia Institute of Technology, my
family, and friends for their encouragement, which played a pivotal
role in my successful completion of this learning experience.
Objectives
01 02 03
Learning Python, data Data cleaning, Creating impactful
analysis, visualization, transformation, visualizations using
machine learning, and organization with NumPy Matplotlib, Seaborn, and
problem-solving. and Pandas libraries. Power BI.

04 05 06
Creating impactful Hands-on experience with Applying academic
visualizations using supervised, unsupervised knowledge to real-world
Matplotlib, Seaborn, and learning, regression, data science challenges
Power BI. classification, clustering. for insights.
Hardware
Requirement
Processor: Intel i5 or higher for efficient
computational performance.
Memory: A minimum of 8GB RAM to handle
data-intensive operations smoothly.
Storage: At least 1GB of free hard disk space,
with additional space for datasets and libraries.
Cache Memory: Minimum of 256KB for quick
data access and processing.
 Software
Requirement
Operating System: Windows, providing a stable and
compatible environment for tools.
Python 3.9 or above: The core programming
language for data manipulation, visualization, and
machine learning.
Jupyter Notebook: For an interactive coding and
data exploration experience.
Visual Studio Code: A robust IDE for Python
development.
Power BI: For creating dynamic dashboards and data
visualizations.
Introduction to Technologies
Python: Python is a high-level, versatile programming language known for its
simplicity, readability, and extensive libraries like NumPy, Pandas, and
Matplotlib, widely used in data science.

NumPy: NumPy is a powerful library for numerical computing, handling arrays

and performing efficient mathematical operations.

Pandas: Pandas is essential for data manipulation, offering DataFrames for

cleaning, filtering, and transformation tasks.
 Numpy
NumPy (Numerical Python) is a fundamental Python library for numerical
computing. It provides support for efficient handling of multidimensional arrays,
matrices, and a variety of mathematical operations. With its performance
optimization, NumPy is faster and more memory-efficient than Python’s built-in
data structures, forming the backbone of many data science and scientific
libraries.

Key Features
Efficiency: NumPy arrays consume less memory and offer faster
computation compared to Python lists.
Extensive Mathematical Functions: NumPy includes a wide range of
functions for linear algebra, statistics, and Fourier transforms.
Support for Multidimensional Arrays: It facilitates operations on
n-dimensional arrays.
 Pandas
Pandas is a powerful Python library for data manipulation and analysis. It provides
versatile data structures like DataFrame and Series, enabling seamless handling
of structured data. Pandas supports data cleaning, filtering, aggregation, and
transformation, making it indispensable for preprocessing in data science
workflows. Its integration with visualization tools enhances data insights
effectively.

Key Features
DataFrame and Series: Powerful data structures for storing data.

Data Cleaning: Handling missing data, removing duplicates, and

transforming data.

Data Aggregation: Grouping and aggregating data using functions like

groupby().
 Data Science
Data Science integrates statistics, computer science, and
machine learning to extract insights from data, involving
collection, cleaning, analysis, and modeling, crucial for
uncovering patterns, predicting trends, and driving organizational
innovation.

Data Collection and Cleaning: Gathering data from various

sources and preparing it by handling missing values,
inconsistencies, and outliers.

Exploratory Data Analysis (EDA): Analyzing data to identify

patterns, relationships, and trends.

Feature Engineering: Creating or selecting relevant features to

improve model performance.

Model Building: Applying machine learning or statistical models to

make predictions or classifications.
Data Visualization
Data Visualization is the graphical representation of
information and data using charts, graphs, and plots. It
simplifies complex datasets, making them easier to
understand and interpret. By translating raw data into
visual formats, visualization aids in recognizing patterns,
trends, and insights, which are critical for informed
decision-making.

It simplifies complex data, reveals patterns, aids

decision-making, ensures effective communication for
stakeholders, and detects anomalies, providing key
insights for informed and accurate decision-making.
Popular Data Visualization Tools

Matplotlib Seaborn Power BI Tableau

A Python library for It is built on It is by Microsoft, Tableau simplifies
static, animated Matplotlib, creates creates interactive interactive
plots, supporting line, attractive statistical dashboards, dashboard creation,
bar, scatter, and graphics and connects data connects diverse
histogram integrates sources, and data sources, and
customization. seamlessly with supports real-time supports complex
Pandas for analysis. monitoring. business intelligence
analyses.
 Power BI
Power BI, developed by Microsoft, is a powerful business analytics tool for
creating interactive reports and dashboards. It enables users to visualize data
from multiple sources and derive actionable insights, making it invaluable for
decision-making.

Key Features
Data Connectivity: Links to various data sources like Excel, SQL databases,
and cloud services.
Interactive Dashboards: Offers dynamic visuals such as bar charts, pie
charts, and geospatial maps.
Real-Time Monitoring: Supports live data updates for timely insights.
Collaboration Tools: Enables sharing of dashboards across teams securely.
 Machine Learning
Machine learning (ML) is a branch of artificial intelligence that enables systems to
learn from data and improve over time without explicit programming. It involves
training algorithms to recognize patterns, make predictions, and perform tasks by
analyzing large datasets. ML can be categorized into three types: supervised
learning, where models are trained with labeled data to predict outcomes;
unsupervised learning, where models discover hidden patterns in unlabeled data;
and reinforcement learning, where agents learn through interactions with an
environment to maximize rewards. ML is used in various applications like
healthcare, finance, marketing, and autonomous systems.
Types of Machine Learning
Supervised Unsupervised Reinforcement
Learning Learning Learning
Supervised learning Unsupervised learning Reinforcement learning
involves training models analyzes unlabeled data involves an agent
on labeled data to to identify patterns, learning through trial
predict outcomes. The structures, or and error by interacting
model learns the relationships without with an environment,
relationship between predefined outputs, receiving rewards or
input and output to commonly used for penalties to maximize
make predictions on clustering, anomaly cumulative rewards
new data. detection, and over time.
dimensionality
reduction.
 Featurres of
Supervised Learning
Labeled Data: Supervised learning relies on labeled datasets, where both
input features and corresponding target outcomes are provided.
Training and Testing Sets: The dataset is divided into training and testing
subsets to evaluate the model’s performance.
Prediction: The primary goal is to predict the target variable based on input
data.
Error Minimization: Models are trained to minimize errors between
predicted and actual outcomes using techniques like gradient descent.
Algorithms: Common algorithms include linear regression, decision trees,
random forests, support vector machines (SVM), and neural networks.
 Classification
Classification is a supervised learning task
where the goal is to predict a categorical
label or class for a given input. The model
learns from labeled data to identify patterns
and then classifies new, unseen data into one
of the predefined categories. Common
applications include spam detection and
image recognition.
 Regression
Regression is a supervised learning
technique used to predict continuous
numerical values based on input features.
The model learns the relationship between
independent variables and the target
variable, making predictions for new data.
Common applications include predicting
house prices, stock market trends, and sales
forecasting.
 Featurres of
Unsupervised Learning
Unlabeled Data: Unsupervised learning works with datasets that don’t have
predefined labels or target outcomes.
Pattern Recognition: The goal is to identify hidden patterns, structures, or
relationships within the data.
Clustering: Grouping similar data points together (e.g., K-means clustering).
Dimensionality Reduction: Reducing the number of features in the dataset
while retaining essential information (e.g., PCA).
No Direct Feedback: Unlike supervised learning, there’s no clear target
variable to guide the learning process.
 Clustering
Clustering is an unsupervised learning
technique used to group similar data points
into clusters based on their features. The
algorithm identifies inherent structures
within the data without predefined labels.
Common clustering algorithms include K-
means, Hierarchical Clustering, and
DBSCAN, used for applications like
customer segmentation and anomaly
detection.
 Association Mining Rule
Association Rule Mining is an unsupervised learning technique used to identify
interesting relationships or patterns in large datasets, typically through "if-then"
rules. It finds frequent itemsets, which are groups of items that frequently appear
together, and generates rules based on these itemsets. A popular algorithm for
association rule mining is Apriori, used in applications like market basket analysis.
This technique helps businesses understand customer behavior, optimize product
placement, and create personalized recommendations in e-commerce or retail.
 Conclusion
In conclusion, this internship provided a comprehensive exposure to data
science and machine learning, allowing me to bridge the gap between
theoretical knowledge and practical application. I gained valuable
experience in Python programming, data analysis using NumPy and Pandas,
and data visualization with tools like Power BI and Matplotlib. I also delved
into machine learning techniques such as supervised, unsupervised, and
reinforcement learning, applying them to real-world problems. This
internship has greatly enhanced my problem-solving skills and has prepared
me for future challenges in the rapidly evolving field of data science and
machine learning.
Thanks!