presentation on AI modeler virtual internship

presentation on AI model

1. IIT Delhi - Virtual Internship
Presented by:- Tanish Malvi
Enrollment no. 0301EC211063
7th Sem ECE Dept.
Submitted to:-
 Prof. Prakash Kumar Singh
ECE Dept.
Madhya Pradesh State Skill Development
Three-Month Online Internship on
Artificial Intelligence Builder By

3. Introduction
Course Introduction:-The Artificial Intelligence Builder course by MPSSDEGB
focuses on foundational and advanced AI techniques.
Internship Objective:-Gain hands-on experience in applying AI and machine
learning to real-world problems.
Activities Overview:-
1. Kaggle exploration and data preprocessing 
2. Implementing CNN and RNN models 
3. Credit Card Fraud Detection project

4. Kaggle
Kaggle is a platform for data science competitions and collaborative projects.
Access Datasets: Download
datasets from Kaggle.
Preprocessing: Clean and
prepare data.
 Model Training: Implement
models using datasets. 
Evaluation: Test and refine
models.

5. Convolutional Neural Networks (CNN)
It is a type of artificial neural network designed for processing structured grid data,
such as images. They are widely used in image recognition, object detection, medical
imaging, and autonomous vehicles.
Importance of CNN 
Feature Extraction
High Accuracy
 Efficiency

6. Recurrent Neural Networks (RNN)
Suppose you want to predict the last word in the text:
 “The clouds are in the ______”.
The most obvious answer to this is the “sky”.
 Consider this sentence: 
I have been staying in Spain for the last 10 years…I can speak fluent ______.
The most suitable answer to this sentence is “Spanish”.
 The word you predict will depend on the previous few words in context. Here, you
need the context of Spain to predict the last word in the text.
It's Deep Learning approach for modelling sequential data. RNNs are a
type of neural network designed to recognize patterns in sequences of
data, such as time series or natural language.

7. Recurrent Neural Networks (RNN)
RNNs are widely used in language
modeling, speech recognition, time 
series prediction, and machine
 translation.

8. Project Title: Credit Card Fraud Detection
Objective of the project : Developed a machine learning model to
accurately identify fraudulent credit card transactions.
Importance of Fraud Detection in the Financial Sector :
Security: Protect financial institutions and customers from significant losses. 
 Trust: Maintain trust in the financial system.
 Financial Stability: Ensure the stability of financial operations.

9. Credit card fraud involves unauthorized transactions that can lead to significant
financial losses. Understanding the types of fraud, such as identity theft and
account takeover, is crucial for effective detection and prevention.
Credit card fraud involves unauthorized transactions that can lead to significant
financial losses. Understanding the types of fraud, such as identity theft and
account takeover, is crucial for effective detection and prevention.

10. Project Details
Preprocessing Steps:- Data Cleaning: Handle missing values and outliers. 
 Feature Engineering: Extract and transform
features.
 Data Normalization: Normalize data to ensure
consistency.
Dataset Used:- Kaggle (Credit Card Fraud Detection Dataset)
Data Splitting: Divide data into training and
testing sets. 
 Training: Train CNN and KNN models. 
 Evaluation: Evaluate model performance using
metrics like accuracy. & Fraud Detection.
Implementation Steps:-

11. Results and Analysis
Summary:
Overall Accuracy: The model achieved
an accuracy of 60-70%. 
Fraud Detection Rate: Successfully
identified 66% of fraudulent
transactions.
False Positive Rate: Reduced false
positives to 37%.

12. Challenges FacedChallenges Faced
Data Quality Issues: Handling missing
values and inconsistencies in the
dataset. 
 Model Performance: Achieving optimal
performance with CNN and KNN
models. 
 Computational Resources: Managing
limited resources for training large
models.
For Solution we referred to YouTube
channels and other websites.

13. Our efforts led to a
significant reduction in fraud
detection time and an
increase in detection
accuracy by over 20%. The
implementation of AI-driven
solutions proved to be a
game changer for financial
security.
Our efforts led to a
significant reduction in fraud
detection time and an
increase in detection
accuracy by over 20%. The
implementation of AI-driven
solutions proved to be a
game changer for financial
security.
Results AchievedResults Achieved

14. ConclusionConclusion
The internship provided hands-on
experience in building and deploying
machine learning models for credit card
fraud detection using Kaggle, CNN and
KNN.
Key takeaways of the internship are
increasing in my overall problem solving
skils, technical skills and practical
knowledge.

15. Thank You!Thank You!
Sincere thanks to the mentors and Madhya Pradesh State Skill
Development for their guidance, support, and the opportunity to
participate in this enriching internship.
A special thank you to my group members Devanand, Sameer &
Priyanshu for their collaboration and contributions throughout the
project. My role involved preprocessing the data and managing all
backend research work.