Update pages/8Model Training.py

pages/8Model Training.py

@@ -20,14 +20,138 @@ if "current_page" not in st.session_state:
 def navigate_to(page_name):
     st.session_state.current_page = page_name
 
-# Main Page
-if st.session_state.current_page == "main":
-    # Reset scroll to top
-    st.query_params.update({})
-
-    # Page Title
-    st.markdown("""
-        <div style="text-align: center; margin-top: 20px;">
-            <h1 style="color: #BB3385;">Model Building</h1>
-        </div>
-    """, unsafe_allow_html=True)
+import streamlit as st
+
+st.set_page_config(page_title="Model Building", layout="wide")
+
+st.markdown("""
+    <h1 style="text-align: center; color: #BB3385;">🛠️ Model Building</h1>
+    <p style="text-align: center; font-size: 18px;">Welcome to one of the most exciting parts of machine learning – teaching the machine how to learn!</p>
+""", unsafe_allow_html=True)
+
+# What is Training?
+st.markdown("## 🤖 So, What is Model Training?")
+st.markdown("""
+Imagine you're a teacher. You give your student (the machine) a bunch of examples and slowly help them learn from it. That’s exactly what model training is.
+
+We give the machine:
+- Some data (like past examples)
+- A method to learn (called an algorithm)
+
+Together, this helps the machine **learn patterns** so it can make decisions or predictions in the future.
+""")
+
+# Who are we training?
+st.markdown("## 👨‍💻 Who are we actually training?")
+st.markdown("""
+We are not training a robot or a human.  
+We are training a **mathematical brain** – called a machine learning model.
+
+You can think of this model like a **blank notebook**.  
+We (programmers) guide it using:
+- The data we have
+- The algorithm we choose
+""")
+
+# What is needed to train
+st.markdown("## 🧠 What does this model need to learn?")
+st.markdown("""
+Only two things:
+1. **Data** – this is like a textbook full of examples  
+2. **Algorithm** – the way the model reads and understands the data  
+
+If the model is not learning properly, and we can’t fix the data, we usually try switching to a better algorithm.
+""")
+
+# Importance of preprocessing
+st.markdown("## 🧹 Why does Preprocessing Matter?")
+st.markdown("""
+Think of this like giving instructions to your student.  
+If you explain in a confusing way, they won’t understand.  
+That’s what happens when we **don’t preprocess the data properly**.
+
+Good learning happens when:
+- Data is cleaned and clear
+- The algorithm matches the task
+""")
+
+# Choosing algorithm type
+st.markdown("## 🤔 Picking the Right Learning Style")
+st.markdown("""
+Before training, we first decide **how the machine should learn**.
+
+We pick from 4 main types:
+- **Supervised** – learning from labeled data (like question + answer)
+- **Unsupervised** – learning without answers (just explore)
+- **Semi-supervised** – mix of both
+- **Reinforcement** – learn by doing (like in games)
+
+Most of the time, we start with **Supervised Learning**.
+""")
+
+# Inside Supervised
+st.markdown("## 🧭 Inside Supervised Learning – Classification vs Regression")
+st.markdown("""
+Now, if you’re using supervised learning, you still need to choose:
+- **Classification** if your answer is a category (like “Spam” or “Not Spam”)
+- **Regression** if your answer is a number (like “House Price = $250,000”)
+
+Choose based on your problem.
+""")
+
+# Data Representation
+st.markdown("## 🧾 How Do We Represent Data to the Model?")
+st.markdown("""
+We write the data in a format the machine understands.  
+
+It usually looks like this:
+
+**D = {(xi, yi)}**
+
+- **xi** is the input (like sepal length, petal width)
+- **yi** is the output (like species of flower)
+
+If yi is a category → it’s **classification**  
+If yi is a number → it’s **regression**
+""")
+
+# Preparing data
+st.markdown("## 📋 Preparing Data Before Training")
+st.markdown("""
+Let’s say we already have cleaned, tabular data. Here’s what we do:
+
+- First, find out the **features** (inputs) and the **target** (output).
+- For example, in the Iris dataset:
+  - Features = sepal length, petal length, etc.
+  - Target = species of flower
+""")
+
+# Train-test split
+st.markdown("## ✂️ Splitting the Data")
+st.markdown("""
+We don’t train on all data.  
+We split it into:
+- **Training Set** – the data we use to teach the model
+- **Testing Set** – the data we use to check how well the model learned
+
+This is like:
+- Studying from textbooks (training)
+- Writing a test paper (testing)
+
+We usually split in ratios like 80:20 or 70:30.
+And remember:
+- No overlap between training and testing data
+- Each data point should have equal chance to be in either group
+""")
+
+# Naming convention
+st.markdown("## 🧾 Naming Things After Split")
+st.markdown("""
+We usually use:
+
+- `X_train`, `y_train` → features and labels for training  
+- `X_test`, `y_test` → features and labels for testing
+""")
+
+# Closing Note
+st.success("🎯 That’s it! You’ve just learned the entire background of how machines get trained. In the next part, we’ll see it in action with a real model.")