Build a Diabetes Dashboard with Python, Streamlit & ML

Name: Build a Diabetes Dashboard with Python, Streamlit & ML
Rating: 5.0 (3 reviews)

A fast-track, project based course covering data science basics, ML, and visualizations.

Created byAnnapoorani Lingeshwaran

Last updated 9/2025

English

English [Auto],

What you'll learn

Build an interactive Streamlit dashboard app in Python from scratch, using real-world diabetes data.
Create insightful data visualizations with Pandas, Matplotlib, and Seaborn to explore health datasets.
Develop and integrate machine learning models (e.g., logistic regression, decision tree) into a deployable web app for diabetes prediction.
Deploy a polished, user-friendly data science project that demonstrates both coding and applied ML skills — perfect for portfolios or job applications.

Course content

5 sections • 30 lectures • 2h 16m total length

Introduction - What we'll create in this course!1:15
Build and deploy a multi-page Streamlit diabetes dashboard, starting with a home page and data visualizations, then train two ML models with interactive feature selection and predictions.
Anaconda Download Instructions5:46
You can lookup Anaconda Download tutorials on youtube to fit your appropriate device; mine is a Windows.
Download Link for Jupyter Files0:11
Opening the Files in Jupyter3:24
NOTE: For mac, there is no 'Extract all' option after right-clicking on the file. Mac users should instead double click on the file for the same result.
Python Crash Course0:13
Explanation of the Dataset3:20
Analyze the diabetes dataset, from a csv to ml-ready data, and explore features like pregnancies, glucose, blood pressure, skin thickness, insulin, bmi, diabetes pedigree function, and age to predict outcomes.

Coding Basic Commands3:42
Uploading Basic Commands to App - Part 16:28
Learn to upload your app design into a Python Streamlit project, create a sidebar with navigation, import data, and run the app locally for deployment.
Uploading Basic Commands to App - Part 24:04
Edit the app home page to include a welcome title, sidebar navigation, and a data frame preview with head and describe; add an expander for the diabetes data set details.

Why Visualize Data?0:47
Histograms and KDEs9:45
Design histograms and KDE plots with Matplotlib and Seaborn, adjusting bins and labels. Layer KDE plots to compare glucose distributions for diabetic versus non-diabetic groups in a Streamlit app.
Adding Histograms to Webapp3:20
Add histograms to the streamlit web app by importing matplotlib and seaborn, creating a button for each graph, and displaying an age histogram with labeled axes.
Adding KDEs to Webapp6:30
Build an interactive KDE plot display by looping through dataframe features, creating per-feature buttons to show distributions by outcome in the web app.
Scatterplots5:59
Learn how to create scatter plots to reveal relationships between glucose and BMI using matplotlib and seaborn, with hue by outcome, and understand subplots and Streamlit integration.
Subplots5:24
Heatmaps and Correlations5:53
Data Cleaning8:19
Learn to clean diabetes data with heat maps that visualize missing values. Replace zeros with NaNs, drop and reassemble columns, and reset indexes to preserve data integrity.
Adding Scatterplots, Correlation Heatmap, and Cleaned Data to App4:08
Add the glucose versus bmi scatterplot, a correlation heat map, and cleaned data to the web app by editing app.py, wiring new visualizations and an expander explanation.

Logistic Regression Theory6:00
Learn logistic regression theory and how it extends linear regression, using a sigmoid to convert outputs into 0-1 probabilities for binary diabetes predictions.
Logistic Regression in Jupyter8:23
Implement logistic regression in a Jupyter notebook to predict diabetes outcomes, including data splitting, scaling, model fitting, and evaluating accuracy.
Extra! In-Depth, Explained: Logistic Regression in Jupyter4:59
Explore the logistic regression workflow in Jupyter, from train-test split and feature scaling to fitting a model and predicting on scaled data, with ones and zeros outcomes.
Explanation of Metrics6:25
Analyze the confusion matrix and classification report to understand precision and recall in diabetes prediction, and interpret sample and real data to evaluate model performance.
Dictionaries Crash Course - Skip if familiar with Python dictionaries1:21
Logistic Regression in Streamlit11:11
NOTE: The line if 'model' in st.session_state should actually be:
if 'model' in st.session_state and 'features' in st.session_state and 'scaler' in st.session_state:
This is to prevent an error if the decision tree classifier is run first. Since we use many of the same variable names, the program gets confused when we look for a scaler that isn't there within the prediction form code, since the model (the decision tree classifier model) is already in session state.

If you would like to avoid these errors all together, change the decision tree classifier variable names in the later lectures. I have kept them the same for simplicity.
Supervised vs. Unsupervised Learning1:23
Decision Tree Classifier Theory3:59
Explore decision tree classifiers in Python, a flowchart-like model that splits data at nodes using a split point, reuses features, and predicts classes for tasks like spam detection.
Decision Trees in Jupyter4:23
Explore building a decision tree classifier in a Jupyter notebook using sklearn, including data preparation with train_test_split, model fitting, prediction, and evaluation via a classification report for diabetes data.
Decision Trees in Streamlit4:14
Build a decision tree classifier in a Streamlit diabetes dashboard. Switch from logistic regression, use train/test split and feature selection, and evaluate with a confusion matrix and classification report.

Requirements

A computer (Windows, Mac, or Linux) with internet access.
Basic familiarity with Python (variables, functions, simple scripts) is helpful but not required — I’ll guide you step by step.
Willingness to install Anaconda (free, beginner-friendly Python distribution) — I’ll walk you through the setup process in detail.
No prior experience with data visualisation, machine learning, or Streamlit is needed — we’ll build everything from scratch.

Description

Are you ready to fast-track your data science journey and build real projects you can proudly showcase? This course is your direct path to becoming a practical, project-oriented data scientist. We've eliminated the endless theory and created a curriculum that gets you hands-on from Day 1. Instead of spending weeks stuck in dry concepts, you’ll learn by doing—working through a complete, project-based curriculum that takes you from raw data all the way to a polished, interactive application.

We'll use the Pima Indians Diabetes Dataset as our guide, a classic challenge that provides the perfect opportunity to master a full, end-to-end data science workflow:

Data exploration & cleaning – Learn how to quickly uncover insights in real-world datasets.
Data visualization – Transform numbers into clear, meaningful charts and graphs.
Machine learning models – Train and evaluate predictive models step-by-step.
Streamlit web apps – Bring your work to life with shareable, interactive dashboards.

A Portfolio Piece That Gets You Noticed

By the end of this course, you won't just "know the concepts"—you’ll have a fully functional data science project to add to your resume, GitHub, or LinkedIn. This isn’t just about earning a certificate; it’s about building a portfolio that proves you have the skills to solve real-world problems. You'll be able to confidently discuss your work, share your code, and showcase a finished product that demonstrates your ability to navigate the entire data science lifecycle. Whether you’re a student, a career-changer, or a busy professional, this fast-track approach ensures you skip the fluff and focus on what really matters: building a skillset through projects.

This course is fast to learn, practical to apply, and built for real results. Stop dreaming about a career in data science and start building your future.

Who this course is for:

Beginner Python developers curious about Data Science
Students interested in conducting medicine/STEM research using ML
Students wanting to learn how to develop an app with Python through Streamlit

Build a Diabetes Dashboard with Python, Streamlit & ML

What you'll learn

Explore related topics

Course content

Introduction6 lectures • 14min

Extracting Basic Insights3 lectures • 14min

Data Visualization9 lectures • 50min

Machine Learning - Logistic Regression & Decision Tree Classifiers10 lectures • 52min

Extra: Overfitting & Deploying the App2 lectures • 6min

Requirements

Description

Who this course is for: