Introduction to Machine Learning: Art of the Possible Flashcards
What is machine learning?
Machine learning (ML) is the process of training computers using mathematical and statistical techniques to recognize patterns in data. Once patterns are identified, ML algorithms create and update models to make increasingly accurate predictions and inferences about future outcomes based on historical and new data. For instance, ML can predict the likelihood of a customer purchasing a product based on previous purchases or past sales history.
What are the key steps in building an ML application?
The key steps in building an ML application are:
Formulate the Problem: Define the problem you want to solve.
Prepare the Data: Gather and preprocess the data needed for training.
Train the Model: Use the data to create and refine a model.
Test the Model: Evaluate the model’s performance on test data.
Deploy the Model: Integrate the model into a production environment
What are some key terms in machine learning?
Model: The output of an ML algorithm trained on a dataset, used for making predictions.
Training: The process of creating a model from past data.
Testing: Measuring the performance of a model on test data.
Deployment: Integrating a model into a production pipeline.
What is the history of Amazon machine learning?
Amazon has over 20 years of experience with machine learning. It uses ML to sell over 4,000 products per minute on Amazon.com and completed its first autonomous Prime Air Delivery in 2016. AWS was recognized as a leader in the Gartner Magic Quadrant for cloud AI developer services in July 2020, with Amazon SageMaker receiving a high rating.
What is the Amazon flywheel approach to machine learning?
The Amazon flywheel, conceptualized by Jeff Bezos, illustrates how investments in key business operations can create a positive feedback loop. By improving customer experience, Amazon attracts more customers, leading to a larger vendor pool and broader product selection. This results in lower prices and cost structures, further enhancing the customer experience. The ML flywheel involves using collected data, predicting outcomes, and continuously improving efficiency and business practices.
How is Amazon using machine learning in its products?
Product Recommendations: Amazon uses ML to provide personalized product recommendations and promotions based on browsing and purchasing data.
Alexa: ML powers voice interactions with Alexa devices using natural language processing (NLP).
Shipping: ML helps Amazon ship 1.6 million packages per day.
How is machine learning helping AWS customers?
AWS machine learning services help customers in various ways:
Amazon Forecast: Delivers accurate forecasts by combining time series data with additional variables.
Amazon Fraud Detector: Identifies fraudulent online activities in milliseconds using ML.
Amazon Personalize: Creates personalized recommendations and marketing experiences.
Amazon Polly: Converts text into lifelike speech.
Amazon Transcribe: Converts speech to text for various applications.
Amazon SageMaker: Provides tools for building, training, and deploying ML models efficiently.
What are some examples of machine learning being used today?
Healthcare: Analyzing clinical data to suggest treatments.
Trucking: Automating logistics.
Ride-Sharing: Updating wait times, demand prediction, and price setting.
Manufacturing: Predicting product defects for cost savings.
Finance: Enabling automated threat intelligence and fraud analysis.
Energy: Improving operations and productivity.
How is the NFL using machine learning?
The NFL uses machine learning and data analytics to enhance the accuracy, speed, and insights provided by its Next Gen Stats (NGS) platform. ML models analyze data from stadiums to improve player health, safety, and fan experiences in real-time.
How can machine learning help me?
Machine learning can:
Make Predictions: Generate forecasts and insights.
Drive Efficiencies: Optimize processes and resource usage.
Enable Automation: Automate repetitive tasks.
Accelerate Decision-Making: Provide timely analysis and support rapid decisions.
What is the potential for machine learning in the future?
The potential of machine learning is vast. It could become an integral part of software engineering, enabling algorithms in low- or no-code environments. Additionally, ML could drive advancements in quantum computing by increasing data processing speeds and accelerating model training.
What percentage of business executives believe AI will impact their revenue and profits?
According to PricewaterhouseCoopers, 48% of US executives believe AI will be a path to growing revenue and profits.
