Machine Learning Concepts and Python Basics

These flashcards cover key machine learning concepts and fundamental Python basics that are essential for understanding the subject.

What is the purpose of Numpy in machine learning?

Numpy is used for arithmetic operations and assumes a grid structure, working with either all numbers or strings.

What is the function of Pandas in machine learning?

Pandas provides data structures that resemble sheets and can handle different types of data, unlike Numpy.

What does the function len() do in Python?

len() is a built-in function used to determine the length or number of items in various data types.

What is the main characteristic of supervised learning?

In supervised learning, the machine is taught by example with provided inputs and outputs.

What type of learning combines labeled and unlabeled data?

Semi-supervised learning combines some defined (labeled) data with additional unlabeled data.

What does unsupervised learning do?

Unsupervised learning identifies patterns and relationships in data without a provided answer key.

What is reinforcement learning in machine learning?

Reinforcement learning involves a machine making decisions based on defined allowed actions and observing outcomes.

What is deep learning?

Deep learning utilizes cognitive computing to process vast amounts of data, improving capabilities in image and sound analysis.

What is the significance of prediction in AI?

Prediction in AI generates missing information from existing data, playing a crucial role in decision making.

How does the concept of cheap technology affect business models?

Cheap technology increases accessibility and alters business models by changing the economics surrounding it.

What is the relationship between AI prediction technology and economics?

AI prediction technology serves as a framework for understanding trade-offs in decision-making within various arenas.

What does the term 'Creative Destruction Lab' refer to?

It is a seed-stage program that increases the probability of success for science-based startups.

What is the role of AI in data categorization?

AI excels at compartmentalizing and categorizing data, completing simple tasks and answering anticipatory questions.

What effect does significant price drop have on behavior?

Significant price drops can change mindsets and behaviors, making previously impossible actions feasible.

Data Literacy,

Understanding and interpreting data effectively.

Causality

Establishing a cause-and-effect relationship

Association,

Identifying correlations between variables.

Observational Studies

Research analyzing existing data without manipulation.

Chocolate and Health

Example of association in health studies.

Death Penalty and Murder Rates

Example of potential causal analysis

Public Data

Freely available datasets for experimentation.

Existing Product Data

User interaction data from current products.

Human-in-the-Loop Systems

Combining automation with human oversight.

Brute Force Collection

Costly data gathering methods for unique datasets.

Purchased Data

Acquired datasets from third-party vendors.

Filtering Impurities,

Managing errors in raw data for quality.

Merging Diverse Data Sources

Integrating datasets from different origins.

Data Labeling

Annotating data for machine learning context.

External Services,

Platforms for scalable data annotation tasks.

Internal Teams,

In-house capabilities for data annotation.

User-Generated Labels

User contributions to data labeling processes.

Annotation Acceleration Tools,

Technologies enhancing data annotation efficiency.

Data Science,

Field combining statistics, machine learning, and domain knowledge.

Collaboration in Data Science

Teamwork essential for solving complex data problems.

Skills of Data Scientists,

Mix of statistics, communication, and visualization skills.

1/5 C's of Data Ethics

Consent

2/5 C’s of data ethics

Clarity

3/5 C’s of data ethics

Consistency

4/5 C’s of data ethics

Control

5/5 C’s of Data ethics

Consequences

IBM Data Estimate

2.5 quintillion bytes of data generated daily.

Prediction in Data Science

Forecasting events based on data analysis.

Productivity Paradox

Technological shifts may delay visible economic benefits.

Big Data

Large datasets requiring responsible usage for impact.

Human Limitations,

Memory and objectivity constraints in data interpretation.

Python Basics,

Fundamental functions for data manipulation in Python.

NumPy

Library for numerical operations in Python

Pandas

Data manipulation library for structured

PyPlot

Matplotlib module for creating visualizations

Seaborn

Statistical data visualization library based on Matplotlib.

Linear Regression

Predicting continuous variables using linear relationships.