cram data science and ai

Mean

Average of all values

most affected by outliers

Median

Middle value of a dataset

Mode

Most frequently occurring value

Range

Difference between maximum and minimum values

Standard deviation

Measure of spread of data from the mean

Variance

Square of standard deviation

Covariance

Measure of how two variables change together

Normal distribution

Bell-shaped symmetric distribution defined by mean and standard deviation

Expected value

Average outcome of a random variable over time

Continuous variable

Variable with infinite possible values such as height

Discrete variable

Countable variable such as number of items

Complement of an event

The probability the event does not occur

Histogram

Chart showing frequency distribution of data

Boxplot

Graph displaying median quartiles and outliers

Scatterplot

Graph showing relationship between two variables

Overfitting

Model learns noise instead of patterns and performs poorly on new data

Underfitting

Model is too simple and doesn’t learn important patterns

Confusion matrix

Table used to evaluate classification models

Precision

How many predicted positives were correct

Recall

How many actual positives were identified correctly

Coefficient of variation

Spread measure useful for comparing datasets with different units

Data cleaning

Correcting or removing inaccurate inconsistent duplicated or missing data

Data normalization

Scaling features into a similar range

Regression

Predicts continuous numerical outcomes

Classification

Predicts categories or labels

SQL

Language used to manage and query relational databases

Primary key

Unique identifier for a database record

Pandas

Python library for data tables and data analysis

NumPy

Python library for numerical computing and arrays

Matplotlib

Python library for data visualization

TensorFlow

Framework for building and training deep learning models

Jupyter Notebook

Interactive tool for writing and running data science code in the browser

Artificial intelligence

Computers performing tasks that normally require human intelligence

Generative AI

AI that creates new content such as text images or audio

AI subfields

Categories including computer vision NLP robotics and human-AI interaction

Large language model

AI system trained to predict and generate text

Supervised learning

Learning based on labeled input and output data

Unsupervised learning

Learning patterns or groups from unlabeled data

Reinforcement learning

Learning through rewards and penalties while interacting with an environment

Training data

Data used to teach a model patterns

Validation data

Data used to tune model hyperparameters

Test data

Data used to evaluate final model performance

Neural network

Model made of interconnected nodes that learn complex patterns

K-means clustering

Unsupervised algorithm that groups similar data points

Deep learning

Machine learning using large multi-layer neural networks

Predicate logic

Logical statements computers can interpret and reason with

Bayesian network

Probabilistic model using nodes and directed edges to represent relationships

Knowledge representation

Storing information in ways computers can understand and reason about

Expert system

AI that uses rules and reasoning to solve problems

Algorithmic bias

Unfair patterns in AI output caused by biased training data

GDPR

European Union law protecting data privacy and data usage transparency

Transparency

Clearly informing users how data is collected processed and used

Data literacy

Ability to read work with analyze and communicate data

Structured data

Data organized in tables and fixed formats

Unstructured data

Data without a fixed format such as images text or emails

Qualitative data

Descriptive non-numeric information

Quantitative data

Numerical measurable information

Streaming data

Real-time continuously arriving data

Batch data

Data processed in groups at scheduled intervals

Metadata

Data that describes other data

Data wrangling

Transforming and preparing raw data for analysis

Data science process

Steps including collecting cleaning analyzing modeling and interpreting data