Machine Learning Overview (PDF)

An Overview  Types of Machine 01 Introduction 07 Learning 02 Activity 08 Supervised Learning Overview of Machine 03 09 Unsupervised Learning Learning...

An Overview  Types of Machine 01 Introduction 07 Learning 02 Activity 08 Supervised Learning Overview of Machine 03 09 Unsupervised Learning Learning Semi – Supervised 04 Types of Data 10 Learning 05 Types of Variables 11 Reinforcement Learning Summary – Machine 06 12 Conclusion Learning  Analyze the data in the dataset to be shown Fill in the blanks based on your understanding Horsepower Fuel Type Drivetrain Price (USD) Model Company 661 Petrol Rear-Wheel 300,000 488 GTB Bugatti 730 Petrol All-Wheel 450,000 Aventador S Lamborghini 640 Petrol All-Wheel 250,000 911 Turbo S Porsche 710 Petrol Rear-Wheel 350,000 720S McLaren 1500 Petrol All-Wheel 3,000,000 Chiron Bugatti 715 Petrol Rear-Wheel 350,000 DBS Superleggera Aston Martin 1600 Petrol Rear-Wheel 4,500,000 Jesko Koenigsegg 791 Petrol Rear-Wheel 3,800,000 Huayra Roadster Pagani 950 Hybrid Rear-Wheel 3,500,000 LaFerrari Ferrari 903 Hybrid Rear-Wheel 2,000,000 P1 McLaren 1600 Petrol Rear-Wheel 4,500,000 Jesko ? 640 Petrol All-Wheel 250,000 911 Turbo S ? 1500 Petrol All-Wheel 3,000,000 Chiron ? 661 Petrol Rear-Wheel 300,000 488 GTB ? A field of artificial intelligence (AI) that focuses on developing algorithms and models that enable computers to learn from and make predictions or decisions based on data.  Works by teaching computers to learn patterns from data 1. Gather data that you want 2. Choose a machine-learning model 5. Refinement and adjustment to understand or predict 3. Train the model for accuracy 4. Test the model Training Data Test Data  Training data is the initial dataset used to train a machine-learning model The model learns patterns and relationships from the training data During the training phase, the model adjusts its internal parameters based on the training data Horsepower Fuel Type Drivetrain Price (USD) Model Company 661 Petrol Rear-Wheel 300,000 488 GTB Bugatti 730 Petrol All-Wheel 450,000 Aventador S Lamborghini 640 Petrol All-Wheel 250,000 911 Turbo S Porsche 710 Petrol Rear-Wheel 350,000 720S McLaren 1500 Petrol All-Wheel 3,000,000 Chiron Bugatti 715 Petrol Rear-Wheel 350,000 DBS Superleggera Aston Martin 1600 Petrol Rear-Wheel 4,500,000 Jesko Koenigsegg 791 Petrol Rear-Wheel 3,800,000 Huayra Roadster Pagani 950 Hybrid Rear-Wheel 3,500,000 LaFerrari Ferrari 903 Hybrid Rear-Wheel 2,000,000 P1 McLaren 1600 Petrol Rear-Wheel 4,500,000 Jesko ? 640 Petrol All-Wheel 250,000 911 Turbo S ? 1500 Petrol All-Wheel 3,000,000 Chiron ? 661 Petrol Rear-Wheel 300,000 488 GTB ?  Test data is a separate dataset not used during training. It is reserved to evaluate the trained machine- learning model's performance. After training, the model is tested with the test data to measure performance metrics like accuracy, precision, and more Horsepower Fuel Type Drivetrain Price (USD) Model Company 661 Petrol Rear-Wheel 300,000 488 GTB Bugatti 730 Petrol All-Wheel 450,000 Aventador S Lamborghini 640 Petrol All-Wheel 250,000 911 Turbo S Porsche 710 Petrol Rear-Wheel 350,000 720S McLaren 1500 Petrol All-Wheel 3,000,000 Chiron Bugatti 715 Petrol Rear-Wheel 350,000 DBS Superleggera Aston Martin 1600 Petrol Rear-Wheel 4,500,000 Jesko Koenigsegg 791 Petrol Rear-Wheel 3,800,000 Huayra Roadster Pagani 950 Hybrid Rear-Wheel 3,500,000 LaFerrari Ferrari 903 Hybrid Rear-Wheel 2,000,000 P1 McLaren 1600 Petrol Rear-Wheel 4,500,000 Jesko ? 640 Petrol All-Wheel 250,000 911 Turbo S ? 1500 Petrol All-Wheel 3,000,000 Chiron ? 661 Petrol Rear-Wheel 300,000 488 GTB ? Independent Variables Dependent Variables  Independent variables provide input data to the model. They are manipulated to observe their effect on the dependent variable. Referred to as predictors, inputs, or features.  Location Size (sq ft) Number of Rooms Price (INR) Mumbai 1000 2 15,000,000 Delhi 1200 3 12,000,000 Bangalore 1500 3 10,000,000 Chennai 1300 3 8,000,000 Hyderabad 1600 4 9,500,000 Pune 1100 2 7,000,000 Kolkata 1400 3 7,500,000 Ahmedabad 1700 4 6,500,000 Jaipur 1200 3 5,500,000 Chandigarh 1800 4 8,000,000 This table is an example of independent and dependent variables  The dependent variable is the variable being tested and measured. It is the outcome that depends on independent variables. Referred to as Target, output, or response variable. The model aims to predict or explain the dependent variable  Location Size (sq ft) Number of Rooms Price (INR) Mumbai 1000 2 15,000,000 Delhi 1200 3 12,000,000 Bangalore 1500 3 10,000,000 Chennai 1300 3 8,000,000 Hyderabad 1600 4 9,500,000 Pune 1100 2 7,000,000 Kolkata 1400 3 7,500,000 Ahmedabad 1700 4 6,500,000 Jaipur 1200 3 5,500,000 Chandigarh 1800 4 8,000,000 Work in pairs Identify the dependent and independent variables from the below example and justify your answer. Work in pairs Price (INR) Kilometres Ran Fuel Type Age of Car (Years) Company Name 500,000 45,000 Petrol 5 Maruti Suzuki 700,000 60,000 Diesel 4 Hyundai 300,000 80,000 Petrol 8 Tata 900,000 30,000 Diesel 3 Mahindra 450,000 70,000 Petrol 6 Honda 650,000 50,000 Diesel 5 Ford 800,000 40,000 Petrol 4 Toyota 350,000 90,000 Diesel 7 Renault 550,000 65,000 Petrol 5 Volkswagen 750,000 55,000 Diesel 4 Nissan Machine learning is a branch of AI focused on creating algorithms and models for computers to learn from data Involves teaching computers to learn and make decisions without explicit programming for each task Commonly used to make predictions or decisions Used for tasks like prediction, classification, recommendation, and optimization Internet Search Social Media Online Shopping Engines Services  Google Search uses machine learning algorithms to improve search results based on user behavior and preferences Internet Search Social Media Online Shopping Engines Services  Facebook and Instagram use machine learning to personalize news feeds, recommend friends, and filter content Internet Search Social Media Online Shopping Engines Services  Amazon and Alibaba use machine learning for product recommendations, customer segmentation, and fraud detection Internet Search Social Media Online Shopping Engines Services  Like humans, computers get better at tasks by learning from examples. After learning the data, computers can make decisions or predictions on new data You don’t need to program every detail; the computer learns from the data provided on its own Supervised Reinforcement Learning Learning Unsupervised Semi-Supervised Learning Learning  Trained on labeled data, where each input has the correct output Understand how inputs relate to outputs to accurately predict new, unseen data. Explain the Process Try to explain the supervised Machine learning process used in the following real life applications Credit Scoring Email Spam Detection Medical Diagnosis Fraud Detection  Learns patterns and structures in data without being given specific instructions or labelled examples Large photo collection on the phone. An unsupervised learning algorithm groups similar photos people, flowers, mountains, etc. Quickly find specific types of photos without manual sorting Explain the Process Try to explain the supervised Machine learning process used in the following real life applications Customer Segmentation Document Clustering Customer Behavior Analysis  Combines both labeled and unlabeled data to train a model. Populated using machine learning techniques Labeled Data: A few emails are marked as "spam" or "not spam.“ Unlabeled Data: Thousands of other emails without labels. The algorithm learns from both to better identify spam emails. Explain the Process Try to explain the supervised Machine learning process used in the following real life applications Speech Recognition Image & Video Analysis Text Classification  The machine learns to make decisions by interacting with an environment Focuses on learning optimal behavior to achieve specific goals through trial and error https://www.youtube.com/watch?v=n2gE7n11h1Y&t=19s Explain the Process Try to explain the supervised Machine learning process used in the following real life applications Game Playing Autonomous vehicles Resource Management  Branch of AI focused on developing algorithms for computers to learn from data and make decisions. Involves training models on large datasets to recognize patterns and generalize to new data. Types of data types in machine learning, training data and Test data Types of variables, Dependent and independent variable  Supervised Learning: Learns from labeled data to predict or classify new data. Unsupervised Learning: Finds patterns in unlabeled data without prior instructions Semi-Supervised Learning: Uses both labeled and unlabeled data and when labeled data is scarce. Reinforcement Learning: Agents learn optimal behaviors through trial and error interactions to maximize rewards. Let’s test our understanding  Machine learning, a pivotal branch of artificial intelligence (AI) Machines use Test data to learn patterns and relationships Increasing or decreasing an independent variable will affect other independent variables  What are the two types of variables in machine learning?  Which Uses labelled data to predict the unlabelled unseen data Clusters unlabelled data by analyzing its patterns Uses both labelled and unlabeled data to train a model  Which aims to predict or classify the test data using what it learned from the training data? Learns by interacting with the environment, using trial and error method Which machine learning type is used in grouping data using similar patterns Happy Learning