Loss Function and Optimization in Supervised Learning

 

Loss Function and Optimization in Supervised Learning

🔹  Overview

In supervised learning, we want to learn a function:

f(x;θ)→yf(x; \theta) \rightarrow y

👉 The key questions:

  • How good is the prediction? → Loss Function
  • How do we improve the model? → Optimization

🔹Loss Function 

🔹Definition

A loss function measures the difference between actual output and predicted output.

L(y,y^)L(y, \hat{y})

Where:

  • yy → true value
  • y^=f(x;θ)\hat{y} = f(x; \theta)→ predicted value

🔹Role of Loss Function

👉 It acts as a guide for learning

  • Quantifies error
  • Provides objective for optimization
  • Helps compare models

🔹Types of Loss Functions

🟦 1. Regression Loss Functions

📌 Mean Squared Error (MSE)

L=(y−y^)2L = (y - \hat{y})^2
  • Penalizes large errors
  • Smooth and differentiable

📌 Mean Absolute Error (MAE)

L=∣y−y^∣L = |y - \hat{y}|
  • Robust to outliers

🟩 2. Classification Loss Functions

📌 0–1 Loss

L={0if correct1if wrongL = \begin{cases} 0 & \text{if correct} \\ 1 & \text{if wrong} \end{cases}

📌 Cross-Entropy Loss (Very Important)

L=−∑ylog⁡(y^)L = - \sum y \log(\hat{y})
  • Used in logistic regression and neural networks

📌 Hinge Loss (SVM)

L=max⁡(0,1−yy^)L = \max(0, 1 - y\hat{y})

🔹 Properties of Good Loss Functions

  • Differentiable
  • Convex (for easy optimization)
  • Reflects real-world cost

🔹Optimization 

🔹 Definition

Optimization is the process of finding model parameters that minimize the loss function.

🔹 Objective Function

min⁡θR(θ)\min_{\theta} R(\theta)

🔹 Gradient Descent

📌 Idea:

Move parameters in direction of steepest decrease of loss

θ=θ−η∇R(θ)\theta = \theta - \eta \nabla R(\theta)

Where:

  • η\eta → learning rate
  • ∇R(θ)\nabla R(\theta) → gradient

🔹 Types of Gradient Descent

🟦 1. Batch Gradient Descent

  • Uses entire dataset

🟩 2. Stochastic Gradient Descent (SGD)

  • Uses one data point

🟨 3. Mini-Batch Gradient Descent

  • Uses small subsets

🔹Learning Rate (Very Important)

  • Too small → slow learning
  • Too large → divergence

🔹 Convergence

👉 Optimization stops when:

  • Loss stops decreasing
  • Gradient ≈ 0

🔹 Local vs Global Minimum

  • Global minimum → best solution
  • Local minimum → suboptimal

🔹 Regularization (Connected to MAP)

Add penalty to avoid overfitting: 

R(θ)+λ∣∣θ∣∣2R(\theta) + \lambda ||\theta||^2

🔹Connection Between Loss and Optimization

👉 Loss defines what to minimize
👉 Optimization defines how to minimize

🔹Example 

📌 Linear Regression

Model:

y^=wx+b\hat{y} = wx + b

Loss:

L=(y−y^)2L = (y - \hat{y})^2

Optimization:

  • Use gradient descent
  • Update w,b

🔹Practical Challenges

  • Non-convex loss (neural networks)
  • Large datasets
  • Vanishing gradients

🔹Key Insights 

  • Learning = optimization problem
  • Choice of loss affects model behavior
  • Efficient optimization is crucial for large-scale data
  • Trade-off between accuracy and computational cost

📝 Summary 

Loss Function:

  • Measures prediction error
  • Defines objective

Optimization:

  • Minimizes loss
  • Finds best parameters

Key Equation:

min⁡θ1n∑L(yi,f(xi))\min_{\theta} \frac{1}{n}\sum L(y_i, f(x_i))

Comments

Post a Comment

Popular posts from this blog

Machine Learning PCCST503 Semester5 KTU CS 2024 Scheme - Dr Binu V P

About Me  - Dr Binu V P Syllabus Machine Learning PCCST503 KTU 2024 Scheme Model Question Paper Machine Learning PCCST503 Semester 5 KTU CS 2024 Scheme Do the programs to better understand the concept. Machine Learning Lab PCCSL508 Module -I Introduction to Machine Learning Introduction to ML Distinguishing Machine Learning from Traditional Programming Machine Learning Paradigms Supervised UnSupervised Semi-Supervised Reinforcement Stages of Machine Learning: From Data Collection to Deployment Parameter Estimation Introduction to Parameter Estimation in Machine Learning Maximum Likelyhood Estimation.(MLE) Maximum A Posteriori Estimation.(MAP) Numerical Problems MLE and MAP Supervised Learning Feature Representation and Problem Formulation Role of loss functions and optimization  Regression Introduction to Regression   Linear regression with one variable  (  lab experiment ) Example Problems- simple linear regression - University Questions Evaluation Metri...
Read more

Introduction to Machine Learning (ML)

  📘 Introduction to Machine Learning (ML) 🔹 What is Machine Learning? Machine Learning is a subfield of Artificial Intelligence that focuses on developing algorithms that allow computers to learn patterns from data and make decisions or predictions without explicit programming . It focuses on developing algorithms and models that allow machines to: Identify patterns in data Make decisions with minimal human intervention Predict future outcomes based on past experiences By continuously learning from new data, machine learning systems become more accurate and efficient over time. 📌 Formal Definition  A computer program is said to learn from experience E with respect to some task T and performance measure P if its performance at task T , as measured by P , improves with experience E . 🔹 Why Machine Learning? Traditional programming: Input + Program → Output Machine Learning paradigm: Input + Output → Model (learned program) 👉 Instead of writing...
Read more

Distinguishing Machine Learning from Traditional Programming

Image
  📘 Distinguishing Machine Learning from Traditional Programming 🔹 1. Fundamental Idea 🧾 Traditional Programming In conventional programming, developers write explicit instructions for the system to follow. The programmer explicitly writes rules/logic System follows deterministic instructions Input data is processed using predefined rules The output is based strictly on those rules 👉 Flow: Input + Program → Output 🤖 Machine Learning In machine learning, the system learns patterns from data instead of being explicitly programmed. Programmer provides  data + learning algorithm The model is trained using data The system  learns rules automatically from data It makes predictions or decisions based on learned patterns 👉 Flow: Input + Output → Learning Algorithm → Model Model + New Input → Predicted Output 🔹 2. Key Difference in Approach Aspect      Traditional Programming Machine Learning Input method Logic creation   ...
Read more