Scribd
Upload
20 views10 pages

ML Web App Presentation

The document presents a web application developed using Streamlit that simplifies machine learning workflows, making them accessible to both technical and non-technical users. It addresses the challenges of traditional ML pipelines by automating data preprocessing, algorithm selection, and result visualization, while integrating various machine learning algorithms. The application has been tested with multiple datasets, demonstrating improved accuracy and efficiency in ML tasks.

Uploaded by

sawleprathamesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
20 views10 pages

ML Web App Presentation

The document presents a web application developed using Streamlit that simplifies machine learning workflows, making them accessible to both technical and non-technical users. It addresses the challenges of traditional ML pipelines by automating data preprocessing, algorithm selection, and result visualization, while integrating various machine learning algorithms. The application has been tested with multiple datasets, demonstrating improved accuracy and efficiency in ML tasks.

Uploaded by

sawleprathamesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 10

Machine Learning

Web App
Using Streamlit
A Streamlined Approach to ML Workflows
With Aegis Digitech
Presented by Prathamesh Sawle
ET-D 19

Guide : Prof. Dr. Swati Shilaskar


Introduction
 Machine learning has become integral to data-driven
decision-making across industries.
 However, the complexity of ML pipelines often acts as
a barrier for non-experts, limiting the adoption of these
technologies.
 Traditional ML workflows require coding expertise,
making it inaccessible to many professionals.
 The project automates these workflows, enabling users
to focus on data insights rather than technical
challenges.
 The application is designed for both technical and non-
technical users, enabling professionals across various
domains to utilize ML effectively without extensive
coding knowledge.
Problem Statement
 Machine learning workflows demand significant
technical expertise, limiting accessibility for non-
experts.
 Repetitive coding for data preprocessing and model
selection is time-consuming and error-prone.
 Existing solutions lack intuitive interfaces for seamless
end-to-end ML pipeline execution.
 Need for a web application to automate preprocessing,
algorithm selection, and result visualization.

 Goal: Simplify ML workflows and democratize access to


robust machine learning tools.
Tools and Technologies
Methodology
Framework Design:
 Built using Streamlit for an interactive and intuitive web
interface.
 Modular architecture to ensure scalability and adaptability for
diverse datasets.

Data Preprocessing:
 Handled missing values using techniques like mean/median
imputation and dropped irrelevant features.
 StandardScaler was applied for feature scaling, while one-hot
and label encoding were used for categorical data.

Algorithm Integration:
 Integrated supervised learning algorithms like Random Forest
and SVM, regression models like Linear Regression, and
clustering techniques like K-Means.
 AutoML implemented for automatic algorithm and
hyperparameter selection based on cross-validation accuracy.
Methodology
Visualization:
 Confusion matrices, PCA-based clustering, and
heatmaps generated for visualizing model results
and correlations.
 Real-time graphs and interactive plots enhance
data interpretability.

Model Evaluation:
 Classification tasks evaluated with metrics like
accuracy, precision, recall, and F1-score.
 Regression models assessed using Mean Squared
Error (MSE) and R-squared values.
 Clustering evaluated using Silhouette Scores.
Experiments
Tested with multiple datasets :

 - Titanic (Classification)
 - Wine Quality (Regression)
 - Mall Customers (Clustering)

 Results include improved accuracy, reduced mean


squared error, and well-defined clusters.
Results

Classification:
 - Titanic Dataset: Random Forest achieved 87%
accuracy.

Regression:
 - Wine Quality Dataset: Linear Regression
achieved R-squared of 0.65.

Clustering:
 - Mall Customers: K-Means achieved a Silhouette
Score of 0.75.
Conclusion

The ML web application successfully simplifies


complex workflows, making machine learning
more accessible to a wider audience.

The integration of AutoML, preprocessing tools,


and interactive visualizations ensures
adaptability and scalability for diverse use cases.
Thank You

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy