VISVESVARAYA TECHNOLOGICAL UNIVERSITY

“JnanaSangama”, Belagavi-590014, Karnataka, India

IMPACT COLLEGE OF ENGINEERING AND APPLIED SCIENCES

Department of Computer Science and Engineering- Data Science

INTERNSHIP PRESENTATION On

“ARTIFICIAL INTELLIGENCE AND

MACHINE LEARNING”
AT
AIROBOSOFT PRODUCTS AND SERVICES
Under the Guidance of :

Submitted By : Dr. Kaipa Sandhya

MONICA K [1IC21CD002] Professor & HOD

Dept. of CS&E- Data Science
 Company Overview (About AiRobosoft)

• AiROBOSOFT, emphasizes the critical importance

of discipline, professionalism, and continuous
learning in shaping the company's culture and
driving its success.

• Discipline at AiROBOSOFT manifests in a

structured approach to project management,
adherence to timelines, and a commitment to quality.

• Professionalism is reflected in AiROBOSOFT's

interactions with clients and partners. The company
maintains a high standard of conduct, characterized
by integrity, transparency, and respect.
 INTRODUCTION
• AI & ML are revolutionizing industries
through automation and intelligent
decision-making.

• This internship at Airobosoft provided

real-world exposure to AI/ML projects.

• Objective: To strengthen technical

expertise and apply machine learning
techniques to solve business problems.
 PRE-REQUISITES

• Technical Knowledge:
 Python Programming
 Basic Data Structures (Lists, Sets, Dictionaries)
 ML Fundamentals (Regression, Classification)

• Software Setup:
 Anaconda
 Jupyter/Colab
 GitHub
 Software Installation
• Download and install Anaconda from
anaconda.com, which includes Python, essential
libraries, and tools for data science. Set
environment variables for Anaconda, especially
for the PATH.

• Jupyter Notebooks is another option for data

science and is included with Anaconda.

• Once everything is installed, you’re all set to

begin coding in Python.
 Python Notebook - 1

• Python syntax is simple, with reserved keywords (like if, else) and
identifiers (user-defined names).
• Use input() for user input and print() for output.
• Common operators include arithmetic (+, -), comparison (==, !=),
logical (and, or), and assignment (=).
• Conditional statements like if and else control program flow, while
loops (for, while) handle repetition.
• Practice with exercises such as printing your name, checking if a
number is positive, or calculating factorials to strengthen your skills.
 Python Notebook - 2
• Python has four key data structures: Lists, Tuples, Sets, and
Dictionaries.
• Lists are ordered and mutable, used for sequences where order matters.
• Tuples are ordered but immutable, ideal for fixed data.
• Sets are unordered and hold unique items, perfect for ensuring no
duplicates.
• Dictionaries store key-value pairs, allowing fast lookups and updates
based on keys.
• Each structure is suited for specific tasks depending on whether you
need order, immutability, uniqueness, or fast access.
 Python Notebook - 3
• Functions in Python are defined using def and can take parameters
(inputs) and return values. They allow for code reusability and
modularity.
• Lambda functions are small, anonymous functions created with the
lambda keyword, typically for short, simple operations.
• File handling allows reading and writing to files using open(), read(),
write(), and close().
• Files can be opened in different modes (read, write, append), and
content can be processed line by line or all at once.
• CSV file handling uses Python's csv module to easily read from and
write to CSV files, which is useful for dealing with structured data in a
 Python Notebook - 4
• Object-Oriented Programming (OOP) is a programming paradigm that
organizes data and behavior into classes and objects.
 A class is a blueprint for creating objects, while an object is an instance of a class.
 Methods are functions defined within a class that describe behaviors or actions of
the objects.
• Inheritance allows a class (child) to inherit properties and methods from
another class (parent), promoting code reusability.
• Polymorphism refers to the ability of different classes to provide different
implementations of the same method, often through method overriding or
method overloading, allowing for flexibility in handling objects.
• Together, these concepts enable the creation of more organized, reusable,
and scalable code.
 Python Notebook - 5

• Modules in Python are files containing Python code (functions,

classes, variables) that can be reused in other programs. You create a
module by saving a Python file with a .py extension, and you can use
it by importing it into other scripts using import module_name.
Packages are collections of related modules grouped together in a
directory with an __init__.py file.

• You can also import standard libraries (built-in modules like math, os,
sys) using import library_name to use pre-built functions and
utilities. This saves time and effort by leveraging Python’s extensive
standard library for tasks like file handling, mathematics, system
 Python Notebook – 6 (NUMPY)
• NumPy (Numerical Python) is a powerful library that
provides support for arrays, enabling efficient
operations on vectors, matrices, and higher-
dimensional data.
• Unlike Python lists, NumPy arrays store elements of
the same type, making them more memory-efficient
and suitable for numerical computations.
• NumPy also includes a wide range of linear algebra
functions such as solving systems of linear equations
and computing Eigenvalues.
• An array is a central data structure that allows for
efficient storage and manipulation of numerical data.
 Python Notebook – 7 (PANDAS)
• Pandas is a powerful Python library for data manipulation,
providing two main structures: Series (1D) and DataFrame (2D).
• Data exploration involves using functions like head() to view
the first few rows, describe() for statistical summaries, and
unique() to find distinct values in columns.
• Data cleaning focuses on handling missing values using fillna()
or dropna() and removing duplicate records with
drop_duplicates(). This ensures data accuracy and consistency.
• Data transformation includes grouping data (groupby()) to
aggregate by categories, sorting (sort_values()) to arrange data
in a specific order, and merging (merge()) DataFrames to
combine related datasets. These operations help organize and
prepare data for deeper analysis.
 Python Notebook – 8 (MATPLOTLIB)
• Matplotlib and Seaborn are powerful Python
libraries for data visualization.
 Matplotlib provides basic plotting functions like line
plots, histograms, and scatter plots
 Seaborn builds on Matplotlib, offering more advanced
visualizations with simpler syntax and enhanced styles.
• Data visualization basics involve creating plots to
better understand data. Common visualizations
include histograms to show distributions, bar
charts to compare categorical data, and pie charts
to illustrate proportions of a whole. Both libraries
allow for customization and styling, making it easy
to create clear, informative visuals for data analysis.
 Python Notebook – 9 (MACHINE
LEARNING)
• Machine Learning is a branch of Artificial Intelligence (AI) that enables computers to learn from data
and improve automatically without being explicitly programmed.
• The process begins with observing data to identify patterns and make better future decisions with
minimal human intervention.
• Solving a machine learning problem generally involves four steps:
 Data preparation
 Training set generation
 Algorithm training
 Model development and Monitoring.
• Data preparation includes selecting, preprocessing, and transforming data to ensure it’s ready
for learning.
• Depending on the task, different algorithms are used, including
 Supervised
 Unsupervised
 Reinforcement
 Semi-supervised learning methods.
 Python Notebook – 10 (LOGISTIC
REGRESSION)
• Logistic Regression is used when the response variable is categorical,
typically binary (0 or 1).
• It models the probability of the outcome using a sigmoid (S-shaped)
curve and applies the logit function to relate predictors linearly to the
log-odds.
• This method ensures outputs are valid probabilities between 0 and 1.
When using categorical predictors, dummy variables are created, and
one is left out to prevent multicollinearity.
• Logistic regression is simple to implement and widely used for
classification tasks like predicting admissions, disease outcomes, or
customer churn.
 WORKING ON SPYDER
• Spyder provides a clean workspace with important panes like the Editor (to
write and edit Python scripts), the IPython Console (to run code
interactively), the Variable Explorer (to view variables in memory), and the
File Explorer (to manage project files).

• You can customize the layout easily based on your workflow: Script
Creation, Running, and Debugging:
• Script Creation: Open Spyder and create a new file (File > New File) to
start writing your Python script.
• Running: Save your script (.py file) and press the green "Run" button or hit
F5 to execute it. Output and results will appear in the IPython Console.
• Debugging: Spyder offers powerful debugging tools — you can set
breakpoints by clicking next to a line number, then start the debugger
(Debug > Debug or press Ctrl + F5). This lets you step through code line-
by-line, inspect variables, and catch errors.
 PROJECT : :

AI PERSONAL ASSISTANT -
JARVIS
 INSTALLING LIBRARIES

• pip install SpeechRecognition: Installs the library to convert speech (audio) into text.
• pip install wikipedia-api: Installs a tool to fetch information from Wikipedia programmatically.
• pip install pyttsx3: Installs a text-to-speech library to make the program speak responses.
• pip install wolframalpha: Installs the client to access WolframAlpha’s powerful knowledge
engine for answering questions.
• pip install instaloader: Installs a tool to download Instagram profile data, posts, and more.
• pip install wikipedia: Installs another simple library to get summaries from Wikipedia (different
from wikipedia-api, simpler use case).
• pip install pyaudio: Installs audio support for microphone input and speaker output, essential for
speech recognition.
• pip install nest_asyncio: Fixes event loop issues in asynchronous Python code, mainly needed
when working with async programs inside environments like Jupyter Notebook.
 ABOUT PROJECT

• The takeCommand() function is designed to capture and process audio input

from the user using the SpeechRecognition library.
• It initializes a speech recognizer (r) and listens for input through the
microphone. Once the audio is captured, the function attempts to convert
the speech into text using Google's speech recognition API with the
language set to English (India).
• If the speech is successfully recognized, it prints the user's statement and
returns it in lowercase.
• If the recognition fails (due to unclear speech or noise), it asks the user to
repeat by calling a speak() function and returns 'none’.
• This function is typically used in voice assistants to interpret user
commands.
• The code in the voice assistant program, where the assistant (named JARVIS) interacts
with the user by listening to voice commands, processing them, and responding
accordingly.
• The program uses a loop to continuously prompt the user with "How can I help you?"
and listens for commands.
• Based on the recognized speech, it performs various actions like opening websites
(YouTube, Google, Gmail), searching Wikipedia, telling the time, fetching news
headlines, answering computational questions using WolframAlpha, and providing
weather updates.
• The assistant also responds to queries about itself and its creator. If the command is
"goodbye," it gracefully shuts down.
• The program handles specific commands using elif blocks and integrates external
services (like WolframAlpha, Wikipedia, and web browsers) for diverse functionalities.
• It also includes error handling to manage situations where the user’s request is unclear or
unsupported.
 SKILLS GAINED

• Python Programming
• Data Analysis
• Machine Learning Basics
• Project Development
• Team Collaboration
 CONCLUSION

• Learning to code, especially using tools like Python and Spyder, is a

transformative journey that takes you from understanding basic
concepts to building real-world applications.
• You start by mastering simple syntax and logic, gradually moving into
handling data, developing models, and solving actual problems.
• With each step, you gain the skills to turn ideas into working projects,
preparing you to tackle real-world challenges with confidence and
creativity.