Building Your First Data Science Project: A Beginner's Step-by-Step Guide to Turn Raw Data into Real Insights

📗 Chapter 7: Building Your First Predictive Model

Train, Test, and Evaluate Your First Real Machine Learning Model in Python

🧠 Introduction

You’ve explored your data, cleaned it, engineered meaningful features, and selected the best ones — now it’s time for the exciting part: building your first predictive model!

In this chapter, you’ll walk through:

• What a predictive model is
• How to split your data into training and testing sets
• Choosing the right algorithm for your task
• Training and evaluating your model
• Interpreting performance metrics
• Improving your model step-by-step

Whether you’re building a classification model to predict survival on the Titanic or a regression model to estimate house prices, this is where your dataset starts providing answers.

🔮 1. What Is a Predictive Model?

A predictive model learns from historical data and makes predictions on new, unseen data.

🔸 Two Most Common Types:

📦 2. Preparing Your Dataset

Make sure:

• Features are numeric (categoricals encoded)
• No missing values
• Scaled/normalized if required

▶ Example Dataset Setup (Titanic-style):

python

X = df.drop('Survived', axis=1)  # Features

y = df['Survived']               # Target

🔀 3. Splitting into Train and Test Sets

Use 80% of the data to train, and 20% to test performance.

python

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

🔍 4. Choosing Your First Algorithm

Start simple:

• Logistic Regression for binary classification
• Linear Regression for predicting continuous values
• Decision Tree for both types

We'll focus on classification using Logistic Regression and Decision Tree.

✅ 5. Logistic Regression (Classification Example)

python

from sklearn.linear_model import LogisticRegression

from sklearn.metrics import accuracy_score

model = LogisticRegression()

model.fit(X_train, y_train)

preds = model.predict(X_test)

print("Accuracy:", accuracy_score(y_test, preds))

🌳 6. Decision Tree Classifier

python

from sklearn.tree import DecisionTreeClassifier

tree = DecisionTreeClassifier(max_depth=4)

tree.fit(X_train, y_train)

tree_preds = tree.predict(X_test)

print("Tree Accuracy:", accuracy_score(y_test, tree_preds))

📈 7. Evaluation Metrics for Classification

▶ Accuracy Score

python

from sklearn.metrics import accuracy_score

print("Accuracy:", accuracy_score(y_test, preds))

▶ Confusion Matrix

python

from sklearn.metrics import confusion_matrix

import seaborn as sns

cm = confusion_matrix(y_test, preds)

sns.heatmap(cm, annot=True, fmt='d')

▶ Precision, Recall, F1-Score

python

from sklearn.metrics import classification_report

print(classification_report(y_test, preds))

📊 8. Evaluation Metrics for Regression

If you’re predicting a numeric value (e.g. house price):

python

from sklearn.linear_model import LinearRegression

from sklearn.metrics import mean_squared_error, r2_score

lr = LinearRegression()

lr.fit(X_train, y_train)

y_pred = lr.predict(X_test)

print("MSE:", mean_squared_error(y_test, y_pred))

print("R2:", r2_score(y_test, y_pred))

🧪 9. Cross-Validation (Optional but Useful)

Get a more stable estimate of model performance.

python

from sklearn.model_selection import cross_val_score

scores = cross_val_score(model, X, y, cv=5, scoring='accuracy')

print("Cross-validated accuracy:", scores.mean())

⚙️ 10. Hyperparameter Tuning

Improve your model by finding the best parameters.

python

from sklearn.model_selection import GridSearchCV

params = {'max_depth': [3, 5, 7, 10]}

grid = GridSearchCV(DecisionTreeClassifier(), param_grid=params, cv=5)

grid.fit(X_train, y_train)

print("Best depth:", grid.best_params_)

🔁 11. Save and Reload Your Model

After training, you can save your model:

python

import joblib

joblib.dump(model, 'my_model.pkl')

model = joblib.load('my_model.pkl')

📋 12. Common Classification Models Overview

🧠 13. Model Selection Tips

📦 14. Summary Table: Model Workflow

✅ Final Code Snippet: Titanic Logistic Regression Example

python

import pandas as pd

from sklearn.model_selection import train_test_split

from sklearn.linear_model import LogisticRegression

from sklearn.metrics import accuracy_score, confusion_matrix, classification_report

import seaborn as sns

import matplotlib.pyplot as plt

# Load dataset

df = pd.read_csv('titanic_clean.csv')

X = df.drop('Survived', axis=1)

y = df['Survived']

# Split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# Train

model = LogisticRegression()

model.fit(X_train, y_train)

# Predict

preds = model.predict(X_test)

# Evaluate

print("Accuracy:", accuracy_score(y_test, preds))

print(classification_report(y_test, preds))

# Confusion Matrix

sns.heatmap(confusion_matrix(y_test, preds), annot=True, fmt='d')

plt.title("Confusion Matrix")

plt.show()