Scikit-learn

Model Deployment with Streamlit and Scikit-learn

Streamlit is a fast, lightweight Python framework used to create interactive web apps for data science and machine learning. It enables quick deployment of Scikit-learn models with a simple UI for real-time predictions.

Key Characteristics

  • No HTML/CSS/JavaScript knowledge required
  • Rapid prototyping of ML interfaces
  • Simple Python scripts render into interactive apps
  • Integrated with Scikit-learn, NumPy, Pandas, and Matplotlib

Basic Rules

  • Save your model using joblib or pickle
  • Use Streamlit widgets (st.text_input, st.slider, etc.) for user input
  • Load and predict with your Scikit-learn model inside the app
  • Run app with streamlit run app.py

Syntax Table

SL NO Task Syntax Example Description
1 Import Streamlit import streamlit as st Loads Streamlit library
2 Load Model model = joblib.load('model.pkl') Load pre-trained model
3 User Input st.text_input("Enter value") Creates input field
4 Predict Result model.predict([inputs]) Predict with model
5 Show Output st.write(f"Prediction: {result}") Displays prediction result

Syntax Explanation

1. Import Streamlit

What is it?
Loads the Streamlit module to build the UI.

Syntax:

import streamlit as st

Explanation:

  • Required to access all Streamlit components
  • Import once at the top of the script

2. Load Model

What is it?
Imports a pre-trained Scikit-learn model for use.

Syntax:

from joblib import load
model = load('model.pkl')

Explanation:

  • Load your model once globally to avoid reloading on each input
  • Make sure to keep the .pkl file in the same folder or provide a valid path

3. User Input

What is it?
Widgets for taking user input in the Streamlit app.

Syntax:

val = st.text_input("Enter feature value")

Explanation:

  • Creates a text box in the UI
  • Accepts numeric or string input depending on use case
  • Can be extended with st.slider, st.selectbox, etc.

4. Predict Result

What is it?
Generates prediction using Scikit-learn model.

Syntax:

prediction = model.predict([[val1, val2, val3]])

Explanation:

  • Input must be reshaped into a 2D array (list of lists)
  • Convert all text input to appropriate data type (float/int)
  • Can wrap in try/except for error handling

5. Show Output

What is it?
Displays prediction result in the Streamlit interface.

Syntax:

st.write("Prediction:", prediction[0])

Explanation:

  • st.write() outputs text, numbers, tables, etc.
  • Used to display dynamic results in app

Real-Life Project: Iris Species Predictor

Project Overview

Build an app that predicts Iris species based on 4 input features using a trained classifier.

Code Example

import streamlit as st
from joblib import load
import numpy as np

# Load model
model = load('iris_model.pkl')

# Title
st.title("Iris Flower Species Predictor")

# Inputs
sepal_length = st.number_input("Sepal Length")
sepal_width = st.number_input("Sepal Width")
petal_length = st.number_input("Petal Length")
petal_width = st.number_input("Petal Width")

# Prediction
if st.button("Predict"):
    inputs = np.array([[sepal_length, sepal_width, petal_length, petal_width]])
    result = model.predict(inputs)
    st.write(f"Prediction: {result[0]}")

Expected Output

  • A simple interactive web UI
  • User inputs feature values and receives model predictions instantly

Common Mistakes to Avoid

  • ❌ Not converting string inputs to float
  • ❌ Model not in same directory or incorrect path
  • ❌ Forgetting to format input into 2D array for .predict()

Further Reading Recommendation

📘 Hands-On Python and Scikit-Learn: A Practical Guide to Machine Learning by Sarful Hassan

🔗 Available on Amazon